Authors: Wieland Morgenstern, Florian Barthel, Anna Hilsmann, Peter Eisert
3D Gaussian Splatting has recently emerged as a highly promising technique for modeling of static 3D scenes. In contrast to Neural Radiance Fields, it utilizes efficient rasterization allowing for very fast rendering at high-quality. However, the storage size is significantly higher, which hinders practical deployment, e.g.~on resource constrained devices. In this paper, we introduce a compact scene representation organizing the parameters of 3D Gaussian Splatting (3DGS) into a 2D grid with local homogeneity, ensuring a drastic reduction in storage requirements without compromising visual quality during rendering. Central to our idea is the explicit exploitation of perceptual redundancies present in natural scenes. In essence, the inherent nature of a scene allows for numerous permutations of Gaussian parameters to equivalently represent it. To this end, we propose a novel highly parallel algorithm that regularly arranges the high-dimensional Gaussian parameters into a 2D grid while preserving their neighborhood structure. During training, we further enforce local smoothness between the sorted parameters in the grid. The uncompressed Gaussians use the same structure as 3DGS, ensuring a seamless integration with established renderers. Our method achieves a reduction factor of 8x to 26x in size for complex scenes with no increase in training time, marking a substantial leap forward in the domain of 3D scene distribution and consumption. Additional information can be found on our project page: https://fraunhoferhhi.github.io/Self-Organizing-Gaussians/
Authors: GuoDong Du, HaoJian Deng, JiaHao Su, Yuan Huang
In this work we address the problem of rain streak removal with RAW images. The general approach is firstly processing RAW data into RGB images and removing rain streak with RGB images. Actually the original information of rain in RAW images is affected by image signal processing (ISP) pipelines including none-linear algorithms, unexpected noise, artifacts and so on. It gains more benefit to directly remove rain in RAW data before being processed into RGB format. To solve this problem, we propose a joint solution for rain removal and RAW processing to obtain clean color images from rainy RAW image. To be specific, we generate rainy RAW data by converting color rain streak into RAW space and design simple but efficient RAW processing algorithms to synthesize both rainy and clean color images. The rainy color images are used as reference to help color corrections. Different backbones show that our method conduct a better result compared with several other state-of-the-art deraining methods focused on color image. In addition, the proposed network generalizes well to other cameras beyond our selected RAW dataset. Finally, we give the result tested on images processed by different ISP pipelines to show the generalization performance of our model is better compared with methods on color images.
Authors: Tao Zhang, Xingye Tian, Yikang Zhou, Shunping Ji, Xuebo Wang, Xin Tao, Yuan Zhang, Pengfei Wan, Zhongyuan Wang, Yu Wu
We present the \textbf{D}ecoupled \textbf{VI}deo \textbf{S}egmentation (DVIS) framework, a novel approach for the challenging task of universal video segmentation, including video instance segmentation (VIS), video semantic segmentation (VSS), and video panoptic segmentation (VPS). Unlike previous methods that model video segmentation in an end-to-end manner, our approach decouples video segmentation into three cascaded sub-tasks: segmentation, tracking, and refinement. This decoupling design allows for simpler and more effective modeling of the spatio-temporal representations of objects, especially in complex scenes and long videos. Accordingly, we introduce two novel components: the referring tracker and the temporal refiner. These components track objects frame by frame and model spatio-temporal representations based on pre-aligned features. To improve the tracking capability of DVIS, we propose a denoising training strategy and introduce contrastive learning, resulting in a more robust framework named DVIS++. Furthermore, we evaluate DVIS++ in various settings, including open vocabulary and using a frozen pre-trained backbone. By integrating CLIP with DVIS++, we present OV-DVIS++, the first open-vocabulary universal video segmentation framework. We conduct extensive experiments on six mainstream benchmarks, including the VIS, VSS, and VPS datasets. Using a unified architecture, DVIS++ significantly outperforms state-of-the-art specialized methods on these benchmarks in both close- and open-vocabulary settings. Code:~\url{https://github.com/zhang-tao-whu/DVIS_Plus}.
Authors: Wenhao Li, Xiu Su, Shan You, Tao Huang, Fei Wang, Chen Qian, Chang Xu
Diffusion models have demonstrated remarkable efficacy in various generative tasks with the predictive prowess of denoising model. Currently, these models employ a uniform denoising approach across all timesteps. However, the inherent variations in noisy latents at each timestep lead to conflicts during training, constraining the potential of diffusion models. To address this challenge, we propose a novel two-stage training strategy termed Step-Adaptive Training. In the initial stage, a base denoising model is trained to encompass all timesteps. Subsequently, we partition the timesteps into distinct groups, fine-tuning the model within each group to achieve specialized denoising capabilities. Recognizing that the difficulties of predicting noise at different timesteps vary, we introduce a diverse model size requirement. We dynamically adjust the model size for each timestep by estimating task difficulty based on its signal-to-noise ratio before fine-tuning. This adjustment is facilitated by a proxy-based structural importance assessment mechanism, enabling precise and efficient pruning of the base denoising model. Our experiments validate the effectiveness of the proposed training strategy, demonstrating an improvement in the FID score on CIFAR10 by over 0.3 while utilizing only 80\% of the computational resources. This innovative approach not only enhances model performance but also significantly reduces computational costs, opening new avenues for the development and application of diffusion models.
Authors: Richard Shaw, Jifei Song, Arthur Moreau, Michal Nazarczuk, Sibi Catley-Chandar, Helisa Dhamo, Eduardo Perez-Pellitero
Novel view synthesis has shown rapid progress recently, with methods capable of producing evermore photo-realistic results. 3D Gaussian Splatting has emerged as a particularly promising method, producing high-quality renderings of static scenes and enabling interactive viewing at real-time frame rates. However, it is currently limited to static scenes only. In this work, we extend 3D Gaussian Splatting to reconstruct dynamic scenes. We model the dynamics of a scene using a tunable MLP, which learns the deformation field from a canonical space to a set of 3D Gaussians per frame. To disentangle the static and dynamic parts of the scene, we learn a tuneable parameter for each Gaussian, which weighs the respective MLP parameters to focus attention on the dynamic parts. This improves the model's ability to capture dynamics in scenes with an imbalance of static to dynamic regions. To handle scenes of arbitrary length whilst maintaining high rendering quality, we introduce an adaptive window sampling strategy to partition the sequence into windows based on the amount of movement in the sequence. We train a separate dynamic Gaussian Splatting model for each window, allowing the canonical representation to change, thus enabling the reconstruction of scenes with significant geometric or topological changes. Temporal consistency is enforced using a fine-tuning step with self-supervising consistency loss on randomly sampled novel views. As a result, our method produces high-quality renderings of general dynamic scenes with competitive quantitative performance, which can be viewed in real-time with our dynamic interactive viewer.
Authors: Haohan Wang, Wei Feng, Yang Lu, Yaoyu Li, Zheng Zhang, Jingjing Lv, Xin Zhu, Junjie Shen, Zhangang Lin, Lixing Bo, Jingping Shao
The state-of-the-art methods for e-commerce product background generation suffer from the inefficiency of designing product-wise prompts when scaling up the production, as well as the ineffectiveness of describing fine-grained styles when customizing personalized backgrounds for some specific brands. To address these obstacles, we integrate the category commonality and personalized style into diffusion models. Concretely, we propose a Category-Wise Generator to enable large-scale background generation for the first time. A unique identifier in the prompt is assigned to each category, whose attention is located on the background by a mask-guided cross attention layer to learn the category-wise style. Furthermore, for products with specific and fine-grained requirements in layout, elements, etc, a Personality-Wise Generator is devised to learn such personalized style directly from a reference image to resolve textual ambiguities, and is trained in a self-supervised manner for more efficient training data usage. To advance research in this field, the first large-scale e-commerce product background generation dataset BG60k is constructed, which covers more than 60k product images from over 2k categories. Experiments demonstrate that our method could generate high-quality backgrounds for different categories, and maintain the personalized background style of reference images. The link to BG60k and codes will be available soon.
Authors: Xinyuan Liu, Lizhi Wang, Lingen Li, Chang Chen, Xue Hu, Fenglong Song, Youliang Yan
Computational spectral imaging is drawing increasing attention owing to the snapshot advantage, and amplitude, phase, and wavelength encoding systems are three types of representative implementations. Fairly comparing and understanding the performance of these systems is essential, but challenging due to the heterogeneity in encoding design. To overcome this limitation, we propose the unified encoding model (UEM) that covers all physical systems using the three encoding types. Specifically, the UEM comprises physical amplitude, physical phase, and physical wavelength encoding models that can be combined with a digital decoding model in a joint encoder-decoder optimization framework to compare the three systems under a unified experimental setup fairly. Furthermore, we extend the UEMs to ideal versions, namely, ideal amplitude, ideal phase, and ideal wavelength encoding models, which are free from physical constraints, to explore the full potential of the three types of computational spectral imaging systems. Finally, we conduct a holistic comparison of the three types of computational spectral imaging systems and provide valuable insights for designing and exploiting these systems in the future.
Authors: Woohyeok Kim, Geonu Kim, Junyong Lee, Seungyong Lee, Seung-Hwan Baek, Sunghyun Cho
RAW images are rarely shared mainly due to its excessive data size compared to their sRGB counterparts obtained by camera ISPs. Learning the forward and inverse processes of camera ISPs has been recently demonstrated, enabling physically-meaningful RAW-level image processing on input sRGB images. However, existing learning-based ISP methods fail to handle the large variations in the ISP processes with respect to camera parameters such as ISO and exposure time, and have limitations when used for various applications. In this paper, we propose ParamISP, a learning-based method for forward and inverse conversion between sRGB and RAW images, that adopts a novel neural-network module to utilize camera parameters, which is dubbed as ParamNet. Given the camera parameters provided in the EXIF data, ParamNet converts them into a feature vector to control the ISP networks. Extensive experiments demonstrate that ParamISP achieve superior RAW and sRGB reconstruction results compared to previous methods and it can be effectively used for a variety of applications such as deblurring dataset synthesis, raw deblurring, HDR reconstruction, and camera-to-camera transfer.
Authors: Tariq Berrada, Jakob Verbeek, Camille Couprie, Karteek Alahari
Semantic image synthesis, i.e., generating images from user-provided semantic label maps, is an important conditional image generation task as it allows to control both the content as well as the spatial layout of generated images. Although diffusion models have pushed the state of the art in generative image modeling, the iterative nature of their inference process makes them computationally demanding. Other approaches such as GANs are more efficient as they only need a single feed-forward pass for generation, but the image quality tends to suffer on large and diverse datasets. In this work, we propose a new class of GAN discriminators for semantic image synthesis that generates highly realistic images by exploiting feature backbone networks pre-trained for tasks such as image classification. We also introduce a new generator architecture with better context modeling and using cross-attention to inject noise into latent variables, leading to more diverse generated images. Our model, which we dub DP-SIMS, achieves state-of-the-art results in terms of image quality and consistency with the input label maps on ADE-20K, COCO-Stuff, and Cityscapes, surpassing recent diffusion models while requiring two orders of magnitude less compute for inference.
Authors: Rongsheng Wang, Qingsong Yao, Haoran Lai, Zhiyang He, Xiaodong Tao, Zihang Jiang, S.Kevin Zhou
Despite significant advancements in medical vision-language pre-training, existing methods have largely overlooked the inherent entity-specific context within radiology reports and the complex cross-modality contextual relationships between text and images. To close this gap, we propose a novel Entity-centered Context-aware Medical Vision-language Pre-training (ECAMP) framework, which is designed to enable a more entity-centered and context-sensitive interpretation of medical data. Utilizing the recent powerful large language model, we distill entity-centered context from medical reports, which enables ECAMP to gain more effective supervision from the text modality. By further pre-training our model with carefully designed entity-aware, context-enhanced masked language modeling and context-guided super-resolution tasks, ECAMP significantly refines the interplay between text and image modalities, leading to an enhanced ability to extract entity-centered contextual features. Besides, our proposed multi-scale context fusion design also improves the semantic integration of both coarse and fine-level image representations, prompting better performance for multi-scale downstream applications. Combining these components leads to significant performance leaps over current state-of-the-art methods and establishes a new standard for cross-modality learning in medical imaging, whose effectiveness is demonstrated by our extensive experiments on various tasks including classification, segmentation, and detection across several public datasets. Code and models are available at https://github.com/ToniChopp/ECAMP.
Authors: Jaesung Rim, Junyong Lee, Heemin Yang, Sunghyun Cho
Mobile cameras, despite their significant advancements, still face low-light challenges due to compact sensors and lenses, leading to longer exposures and motion blur. Traditional solutions like blind deconvolution and learning-based methods often fall short in handling ill-posedness of the deblurring problem. To address this, we propose a novel deblurring framework for multi-camera smartphones, utilizing a hybrid imaging technique. We simultaneously capture a long exposure wide-angle image and ultra-wide burst images from a smartphone, and use the sharp burst to estimate blur kernels for deblurring the wide-angle image. For learning and evaluation of our network, we introduce the HCBlur dataset, which includes pairs of blurry wide-angle and sharp ultra-wide burst images, and their sharp wide-angle counterparts. We extensively evaluate our method, and the result shows the state-of-the-art quality.
Authors: Xin Wang, Lizhi Wang, Xiangtian Ma, Maoqing Zhang, Lin Zhu, Hua Huang
Dual-Camera Compressed Hyperspectral Imaging (DCCHI) offers the capability to reconstruct 3D Hyperspectral Image (HSI) by fusing compressive and Panchromatic (PAN) image, which has shown great potential for snapshot hyperspectral imaging in practice. In this paper, we introduce a novel DCCHI reconstruction network, the Intra-Inter Similarity Exploiting Transformer (In2SET). Our key insight is to make full use of the PAN image to assist the reconstruction. To this end, we propose using the intra-similarity within the PAN image as a proxy for approximating the intra-similarity in the original HSI, thereby offering an enhanced content prior for more accurate HSI reconstruction. Furthermore, we aim to align the features from the underlying HSI with those of the PAN image, maintaining semantic consistency and introducing new contextual information for the reconstruction process. By integrating In2SET into a PAN-guided unrolling framework, our method substantially enhances the spatial-spectral fidelity and detail of the reconstructed images, providing a more comprehensive and accurate depiction of the scene. Extensive experiments conducted on both real and simulated datasets demonstrate that our approach consistently outperforms existing state-of-the-art methods in terms of reconstruction quality and computational complexity. Code will be released.
Authors: Weijia Mao, Yan-Pei Cao, Jia-Wei Liu, Zhongcong Xu, Mike Zheng Shou
We introduce ShowRoom3D, a three-stage approach for generating high-quality 3D room-scale scenes from texts. Previous methods using 2D diffusion priors to optimize neural radiance fields for generating room-scale scenes have shown unsatisfactory quality. This is primarily attributed to the limitations of 2D priors lacking 3D awareness and constraints in the training methodology. In this paper, we utilize a 3D diffusion prior, MVDiffusion, to optimize the 3D room-scale scene. Our contributions are in two aspects. Firstly, we propose a progressive view selection process to optimize NeRF. This involves dividing the training process into three stages, gradually expanding the camera sampling scope. Secondly, we propose the pose transformation method in the second stage. It will ensure MVDiffusion provide the accurate view guidance. As a result, ShowRoom3D enables the generation of rooms with improved structural integrity, enhanced clarity from any view, reduced content repetition, and higher consistency across different perspectives. Extensive experiments demonstrate that our method, significantly outperforms state-of-the-art approaches by a large margin in terms of user study.
Authors: Zinuo You, Andreas Geiger, Anpei Chen
We present NeLF-Pro, a novel representation for modeling and reconstructing light fields in diverse natural scenes that vary in extend and spatial granularity. In contrast to previous fast reconstruction methods that represent the 3D scene globally, we model the light field of a scene as a set of local light field feature probes, parameterized with position and multi-channel 2D feature maps. Our central idea is to bake the scene's light field into spatially varying learnable representations and to query point features by weighted blending of probes close to the camera - allowing for mipmap representation and rendering. We introduce a novel vector-matrix-matrix (VMM) factorization technique that effectively represents the light field feature probes as products of core factors (i.e., VM) shared among local feature probes, and a basis factor (i.e., M) - efficiently encoding internal relationships and patterns within the scene.Experimentally, we demonstrate that NeLF-Pro significantly boosts the performance of feature grid-based representations, and achieves fast reconstruction with better rendering quality while maintaining compact modeling.
Authors: Yunchuan Ma, Chang Teng, Yuankai Qi, Guorong Li, Laiyu Qing, Qi Wu, Qingming Huang
Describing video content according to users' needs is a long-held goal. Although existing video captioning methods have made significant progress, the generated captions may not focus on the entity that users are particularly interested in. To address this problem, we propose a new video captioning task, subject-oriented video captioning, which allows users to specify the describing target via a bounding box. To support this task, we construct two subject-oriented video captioning datasets based on two widely used video captioning datasets: MSVD and MSRVTT, by annotating subjects in each video for each caption. These datasets pave the way for future technique development. As the first attempt, we evaluate four state-of-the-art general video captioning models, and have observed a large performance drop. We then explore several strategies to enable them to describe the desired target. Experimental results show obvious improvement, but there is still a large room for further exploration in this field.
Authors: Junru Lin, Asen Nachkov, Songyou Peng, Luc Van Gool, Danda Pani Paudel
The opacity of rigid 3D scenes with opaque surfaces is considered to be of a binary type. However, we observed that this property is not followed by the existing RGB-only NeRF-SLAM. Therefore, we are motivated to introduce this prior into the RGB-only NeRF-SLAM pipeline. Unfortunately, the optimization through the volumetric rendering function does not facilitate easy integration of the desired prior. Instead, we observed that the opacity of ternary-type (TT) is well supported. In this work, we study why ternary-type opacity is well-suited and desired for the task at hand. In particular, we provide theoretical insights into the process of jointly optimizing radiance and opacity through the volumetric rendering process. Through exhaustive experiments on benchmark datasets, we validate our claim and provide insights into the optimization process, which we believe will unleash the potential of RGB-only NeRF-SLAM. To foster this line of research, we also propose a simple yet novel visual odometry scheme that uses a hybrid combination of volumetric and warping-based image renderings. More specifically, the proposed hybrid odometry (HO) additionally uses image warping-based coarse odometry, leading up to an order of magnitude final speed-up. Furthermore, we show that the proposed TT and HO well complement each other, offering state-of-the-art results on benchmark datasets in terms of both speed and accuracy.
Authors: David Pujol-Perich, Albert Clapés, Sergio Escalera
Temporal Action Localization (TAL) is a complex task that poses relevant challenges, particularly when attempting to generalize on new -- unseen -- domains in real-world applications. These scenarios, despite realistic, are often neglected in the literature, exposing these solutions to important performance degradation. In this work, we tackle this issue by introducing, for the first time, an approach for Unsupervised Domain Adaptation (UDA) in sparse TAL, which we refer to as Semantic Adversarial unsupervised Domain Adaptation (SADA). Our contribution is threefold: (1) we pioneer the development of a domain adaptation model that operates on realistic sparse action detection benchmarks; (2) we tackle the limitations of global-distribution alignment techniques by introducing a novel adversarial loss that is sensitive to local class distributions, ensuring finer-grained adaptation; and (3) we present a novel experimental setup, based on EpicKitchens100, that evaluates multiple types of domain shifts in a comprehensive manner. Our experimental results indicate that SADA improves the adaptation across domains when compared to fully supervised state-of-the-art and alternative UDA methods, attaining a relative performance boost of up to 14%.
Authors: Xin Xu, Jinman Park, Paul Fieguth
Single-image super-resolution (SISR) has seen significant advancements through the integration of deep learning. However, the substantial computational and memory requirements of existing methods often limit their practical application. This paper introduces a new Efficient Pyramid Network (EPNet) that harmoniously merges an Edge Split Pyramid Module (ESPM) with a Panoramic Feature Extraction Module (PFEM) to overcome the limitations of existing methods, particularly in terms of computational efficiency. The ESPM applies a pyramid-based channel separation strategy, boosting feature extraction while maintaining computational efficiency. The PFEM, a novel fusion of CNN and Transformer structures, enables the concurrent extraction of local and global features, thereby providing a panoramic view of the image landscape. Our architecture integrates the PFEM in a manner that facilitates the streamlined exchange of feature information and allows for the further refinement of image texture details. Experimental results indicate that our model outperforms existing state-of-the-art methods in image resolution quality, while considerably decreasing computational and memory costs. This research contributes to the ongoing evolution of efficient and practical SISR methodologies, bearing broader implications for the field of computer vision.
Authors: Madhuri Nagare, Gregery T. Buzzard, Charles A. Bouman
Deep neural networks (DNN) are commonly used to denoise and sharpen X-ray computed tomography (CT) images with the goal of reducing patient X-ray dosage while maintaining reconstruction quality. However, naive application of DNN-based methods can result in image texture that is undesirable in clinical applications. Alternatively, generative adversarial network (GAN) based methods can produce appropriate texture, but naive application of GANs can introduce inaccurate or even unreal image detail. In this paper, we propose a texture matching generative adversarial network (TMGAN) that enhances CT images while generating an image texture that can be matched to a target texture. We use parallel generators to separate anatomical features from the generated texture, which allows the GAN to be trained to match the desired texture without directly affecting the underlying CT image. We demonstrate that TMGAN generates enhanced image quality while also producing image texture that is desirable for clinical application.
Authors: Tonmoy Hossain, Jian Wang, Miaomiao Zhang
Geometric transformations have been widely used to augment the size of training images. Existing methods often assume a unimodal distribution of the underlying transformations between images, which limits their power when data with multimodal distributions occur. In this paper, we propose a novel model, Multimodal Geometric Augmentation (MGAug), that for the first time generates augmenting transformations in a multimodal latent space of geometric deformations. To achieve this, we first develop a deep network that embeds the learning of latent geometric spaces of diffeomorphic transformations (a.k.a. diffeomorphisms) in a variational autoencoder (VAE). A mixture of multivariate Gaussians is formulated in the tangent space of diffeomorphisms and serves as a prior to approximate the hidden distribution of image transformations. We then augment the original training dataset by deforming images using randomly sampled transformations from the learned multimodal latent space of VAE. To validate the efficiency of our model, we jointly learn the augmentation strategy with two distinct domain-specific tasks: multi-class classification on 2D synthetic datasets and segmentation on real 3D brain magnetic resonance images (MRIs). We also compare MGAug with state-of-the-art transformation-based image augmentation algorithms. Experimental results show that our proposed approach outperforms all baselines by significantly improved prediction accuracy. Our code is publicly available at https://github.com/tonmoy-hossain/MGAug.
Authors: Jiawei Yao, Xiaochao Pan, Tong Wu, Xiaofeng Zhang
Detecting lane lines from sensors is becoming an increasingly significant part of autonomous driving systems. However, less development has been made on high-definition lane-level mapping based on aerial images, which could automatically build and update offline maps for auto-driving systems. To this end, our work focuses on extracting fine-level detailed lane lines together with their topological structures. This task is challenging since it requires large amounts of data covering different lane types, terrain and regions. In this paper, we introduce for the first time a large-scale aerial image dataset built for lane detection, with high-quality polyline lane annotations on high-resolution images of around 80 kilometers of road. Moreover, we developed a baseline deep learning lane detection method from aerial images, called AerialLaneNet, consisting of two stages. The first stage is to produce coarse-grained results at point level, and the second stage exploits the coarse-grained results and feature to perform the vertex-matching task, producing fine-grained lanes with topology. The experiments show our approach achieves significant improvement compared with the state-of-the-art methods on our new dataset. Our code and new dataset are available at https://github.com/Jiawei-Yao0812/AerialLaneNet.
Authors: Sudharshan Suresh, Haozhi Qi, Tingfan Wu, Taosha Fan, Luis Pineda, Mike Lambeta, Jitendra Malik, Mrinal Kalakrishnan, Roberto Calandra, Michael Kaess, Joseph Ortiz, Mustafa Mukadam
To achieve human-level dexterity, robots must infer spatial awareness from multimodal sensing to reason over contact interactions. During in-hand manipulation of novel objects, such spatial awareness involves estimating the object's pose and shape. The status quo for in-hand perception primarily employs vision, and restricts to tracking a priori known objects. Moreover, visual occlusion of objects in-hand is imminent during manipulation, preventing current systems to push beyond tasks without occlusion. We combine vision and touch sensing on a multi-fingered hand to estimate an object's pose and shape during in-hand manipulation. Our method, NeuralFeels, encodes object geometry by learning a neural field online and jointly tracks it by optimizing a pose graph problem. We study multimodal in-hand perception in simulation and the real-world, interacting with different objects via a proprioception-driven policy. Our experiments show final reconstruction F-scores of $81$% and average pose drifts of $4.7\,\text{mm}$, further reduced to $2.3\,\text{mm}$ with known CAD models. Additionally, we observe that under heavy visual occlusion we can achieve up to $94$% improvements in tracking compared to vision-only methods. Our results demonstrate that touch, at the very least, refines and, at the very best, disambiguates visual estimates during in-hand manipulation. We release our evaluation dataset of 70 experiments, FeelSight, as a step towards benchmarking in this domain. Our neural representation driven by multimodal sensing can serve as a perception backbone towards advancing robot dexterity. Videos can be found on our project website https://suddhu.github.io/neural-feels/
Authors: Jens Naumann, Binbin Xu, Stefan Leutenegger, Xingxing Zuo
We introduce a novel monocular visual odometry (VO) system, NeRF-VO, that integrates learning-based sparse visual odometry for low-latency camera tracking and a neural radiance scene representation for sophisticated dense reconstruction and novel view synthesis. Our system initializes camera poses using sparse visual odometry and obtains view-dependent dense geometry priors from a monocular depth prediction network. We harmonize the scale of poses and dense geometry, treating them as supervisory cues to train a neural implicit scene representation. NeRF-VO demonstrates exceptional performance in both photometric and geometric fidelity of the scene representation by jointly optimizing a sliding window of keyframed poses and the underlying dense geometry, which is accomplished through training the radiance field with volume rendering. We surpass state-of-the-art methods in pose estimation accuracy, novel view synthesis fidelity, and dense reconstruction quality across a variety of synthetic and real-world datasets, while achieving a higher camera tracking frequency and consuming less GPU memory.
Authors: Demircan Tas, Mine Özkar
In three-dimensional models obtained by photogrammetry of existing structures, all of the shapes that the eye can select cannot always find their equivalents in the geometric components of the model. However, the matching of meaningful parts and assemblages with the records acquired with rapid and detailed documentation methods will provide an advantage for the creation of information models of existing structures. While aiming to produce answers to this problem and in order to overcome the difficulties of pattern recognition in three-dimensional models, we used two-dimensional samples obtained by projection. Processing techniques such as ambient occlusion, curvature and normal maps are commonly used in modern computer graphics applications that enable the representation of three-dimensional surface properties in two-dimensional data sets. The method we propose is based on the recognition of patterns through these mappings instead of the usual light-based visualization. The first stage of the application is photogrammetric capture of a few examples of Zeugma mosaics and three-dimensional digital modeling of a set of Seljuk era brick walls based on knowledge obtained through architectural history literature. The second stage covers the creation of digital models byprocessing the surface representation obtained from this data using Alice Vision, OpenCV-Python, and Autodesk Maya to include information on aspects of the making of the walls. What is envisioned for the next stages is that the mapping data contributes and supports the knowledge for rule-based design and making processesof cultural heritage.
Authors: David Nakath, Xiangyu Weng, Mengkun She, Kevin Köser
When created faithfully from real-world data, Digital 3D representations of objects can be useful for human or computer-assisted analysis. Such models can also serve for generating training data for machine learning approaches in settings where data is difficult to obtain or where too few training data exists, e.g. by providing novel views or images in varying conditions. While the vast amount of visual 3D reconstruction approaches focus on non-physical models, textured object surfaces or shapes, in this contribution we propose a volumetric reconstruction approach that obtains a physical model including the interior of partially translucent objects such as plankton or insects. Our technique photographs the object under different poses in front of a bright white light source and computes absorption and scattering per voxel. It can be interpreted as visual tomography that we solve by inverse raytracing. We additionally suggest a method to convert non-physical NeRF media into a physically-based volumetric grid for initialization and illustrate the usefulness of the approach using two real-world plankton validation sets, the lab-scanned models being finally also relighted and virtually submerged in a scenario with augmented medium and illumination conditions. Please visit the project homepage at www.marine.informatik.uni-kiel.de/go/vito
Authors: Lixu Wang, Chenxi Liu, Junfeng Guo, Jiahua Dong, Xiao Wang, Heng Huang, Qi Zhu
In a privacy-focused era, Federated Learning (FL) has emerged as a promising machine learning technique. However, most existing FL studies assume that the data distribution remains nearly fixed over time, while real-world scenarios often involve dynamic and continual changes. To equip FL systems with continual model evolution capabilities, we focus on an important problem called Federated Continual Novel Class Learning (FedCN) in this work. The biggest challenge in FedCN is to merge and align novel classes that are discovered and learned by different clients without compromising privacy. To address this, we propose a Global Alignment Learning (GAL) framework that can accurately estimate the global novel class number and provide effective guidance for local training from a global perspective, all while maintaining privacy protection. Specifically, GAL first locates high-density regions in the representation space through a bi-level clustering mechanism to estimate the novel class number, with which the global prototypes corresponding to novel classes can be constructed. Then, GAL uses a novel semantic weighted loss to capture all possible correlations between these prototypes and the training data for mitigating the impact of pseudo-label noise and data heterogeneity. Extensive experiments on various datasets demonstrate GAL's superior performance over state-of-the-art novel class discovery methods. In particular, GAL achieves significant improvements in novel-class performance, increasing the accuracy by 5.1% to 10.6% in the case of one novel class learning stage and by 7.8% to 17.9% in the case of two novel class learning stages, without sacrificing known-class performance. Moreover, GAL is shown to be effective in equipping a variety of different mainstream FL algorithms with novel class discovery and learning capability, highlighting its potential for many real-world applications.
Authors: Bingbing Wen, Zhengyuan Yang, Jianfeng Wang, Zhe Gan, Bill Howe, Lijuan Wang
In this paper, we build a visual dialogue dataset, named InfoVisDial, which provides rich informative answers in each round even with external knowledge related to the visual content. Different from existing datasets where the answer is compact and short, InfoVisDial contains long free-form answers with rich information in each round of dialogue. For effective data collection, the key idea is to bridge the large-scale multimodal model (e.g., GIT) and the language models (e.g., GPT-3). GIT can describe the image content even with scene text, while GPT-3 can generate informative dialogue based on the image description and appropriate prompting techniques. With such automatic pipeline, we can readily generate informative visual dialogue data at scale. Then, we ask human annotators to rate the generated dialogues to filter the low-quality conversations.Human analyses show that InfoVisDial covers informative and diverse dialogue topics: $54.4\%$ of the dialogue rounds are related to image scene texts, and $36.7\%$ require external knowledge. Each round's answer is also long and open-ended: $87.3\%$ of answers are unique with an average length of $8.9$, compared with $27.37\%$ and $2.9$ in VisDial. Last, we propose a strong baseline by adapting the GIT model for the visual dialogue task and fine-tune the model on InfoVisDial. Hopefully, our work can motivate more effort on this direction.
Authors: Youssef Mourchid, Marc Donias, Yannick Berthoumieu, Mohamed Najim
This paper addresses the automatic colorization problem, which converts a gray-scale image to a colorized one. Recent deep-learning approaches can colorize automatically grayscale images. However, when it comes to different scenes which contain distinct color styles, it is difficult to accurately capture the color characteristics. In this work, we propose a fully automatic colorization approach based on Symmetric Positive Definite (SPD) Manifold Learning with a generative adversarial network (SPDGAN) that improves the quality of the colorization results. Our SPDGAN model establishes an adversarial game between two discriminators and a generator. The latter is based on ResNet architecture with few alterations. Its goal is to generate fake colorized images without losing color information across layers through residual connections. Then, we employ two discriminators from different domains. The first one is devoted to the image pixel domain, while the second one is to the Riemann manifold domain which helps to avoid color misalignment. Extensive experiments are conducted on the Places365 and COCO-stuff databases to test the effect of each component of our SPDGAN. In addition, quantitative and qualitative comparisons with state-of-the-art methods demonstrate the effectiveness of our model by achieving more realistic colorized images with less artifacts visually, and good results of PSNR, SSIM, and FID values.
Authors: Youssef Mourchid, Rim Slama
Accurate assessment of patient actions plays a crucial role in healthcare as it contributes significantly to disease progression monitoring and treatment effectiveness. However, traditional approaches to assess patient actions often rely on manual observation and scoring, which are subjective and time-consuming. In this paper, we propose an automated approach for patient action assessment using a Multi-Residual Spatio Temporal Graph Network (MR-STGN) that incorporates both angular and positional 3D skeletons. The MR-STGN is specifically designed to capture the spatio-temporal dynamics of patient actions. It achieves this by integrating information from multiple residual layers, with each layer extracting features at distinct levels of abstraction. Furthermore, we integrate an attention fusion mechanism into the network, which facilitates the adaptive weighting of various features. This empowers the model to concentrate on the most pertinent aspects of the patient's movements, offering precise instructions regarding specific body parts or movements that require attention. Ablation studies are conducted to analyze the impact of individual components within the proposed model. We evaluate our model on the UI-PRMD dataset demonstrating its performance in accurately predicting real-time patient action scores, surpassing state-of-the-art methods.
Authors: Jingdong Zhang, Jiayuan Fan, Peng Ye, Bo Zhang, Hancheng Ye, Baopu Li, Yancheng Cai, Tao Chen
Existing works generally adopt the encoder-decoder structure for Multi-task Dense Prediction, where the encoder extracts the task-generic features, and multiple decoders generate task-specific features for predictions. We observe that low-level representations with rich details and high-level representations with abundant task information are not both involved in the multi-task interaction process. Additionally, low-quality and low-efficiency issues also exist in current multi-task learning architectures. In this work, we propose to learn a comprehensive intermediate feature globally from both task-generic and task-specific features, we reveal an important fact that this intermediate feature, namely the bridge feature, is a good solution to the above issues. Based on this, we propose a novel Bridge-Feature-Centirc Interaction (BRFI) method. A Bridge Feature Extractor (BFE) is designed for the generation of strong bridge features and Task Pattern Propagation (TPP) is applied to ensure high-quality task interaction participants. Then a Task-Feature Refiner (TFR) is developed to refine final task predictions with the well-learned knowledge from the bridge features. Extensive experiments are conducted on NYUD-v2 and PASCAL Context benchmarks, and the superior performance shows the proposed architecture is effective and powerful in promoting different dense prediction tasks simultaneously.
Authors: Minh-Quan Viet Bui, Jongmin Park, Jihyong Oh, Munchurl Kim
Video view synthesis, allowing for the creation of visually appealing frames from arbitrary viewpoints and times, offers immersive viewing experiences. Neural radiance fields, particularly NeRF, initially developed for static scenes, have spurred the creation of various methods for video view synthesis. However, the challenge for video view synthesis arises from motion blur, a consequence of object or camera movement during exposure, which hinders the precise synthesis of sharp spatio-temporal views. In response, we propose a novel dynamic deblurring NeRF framework for blurry monocular video, called DyBluRF, consisting of an Interleave Ray Refinement (IRR) stage and a Motion Decomposition-based Deblurring (MDD) stage. Our DyBluRF is the first that addresses and handles the novel view synthesis for blurry monocular video. The IRR stage jointly reconstructs dynamic 3D scenes and refines the inaccurate camera pose information to combat imprecise pose information extracted from the given blurry frames. The MDD stage is a novel incremental latent sharp-rays prediction (ILSP) approach for the blurry monocular video frames by decomposing the latent sharp rays into global camera motion and local object motion components. Extensive experimental results demonstrate that our DyBluRF outperforms qualitatively and quantitatively the very recent state-of-the-art methods. Our project page including source codes and pretrained model are publicly available at https://kaist-viclab.github.io/dyblurf-site/.
Authors: Benjamin Billot, Daniel Moyer, Neel Dey, Malte Hoffmann, Esra Abaci Turk, Borjan Gagoski, Ellen Grant, Polina Golland
Rigid motion tracking is paramount in many medical imaging applications where movements need to be detected, corrected, or accounted for. Modern strategies rely on convolutional neural networks (CNN) and pose this problem as rigid registration. Yet, CNNs do not exploit natural symmetries in this task, as they are equivariant to translations (their outputs shift with their inputs) but not to rotations. Here we propose EquiTrack, the first method that uses recent steerable SE(3)-equivariant CNNs (E-CNN) for motion tracking. While steerable E-CNNs can extract corresponding features across different poses, testing them on noisy medical images reveals that they do not have enough learning capacity to learn noise invariance. Thus, we introduce a hybrid architecture that pairs a denoiser with an E-CNN to decouple the processing of anatomically irrelevant intensity features from the extraction of equivariant spatial features. Rigid transforms are then estimated in closed-form. EquiTrack outperforms state-of-the-art learning and optimisation methods for motion tracking in adult brain MRI and fetal MRI time series. Our code is available at github.com/BBillot/equitrack.
Authors: Hai Zhang, Chunwei Wu, Guitao Cao, Hailing Wang, Wenming Cao
Editing real images authentically while also achieving cross-domain editing remains a challenge. Recent studies have focused on converting real images into latent codes and accomplishing image editing by manipulating these codes. However, merely manipulating the latent codes would constrain the edited images to the generator's image domain, hindering the attainment of diverse editing goals. In response, we propose an innovative image editing method called HyperEditor, which utilizes weight factors generated by hypernetworks to reassign the weights of the pre-trained StyleGAN2's generator. Guided by CLIP's cross-modal image-text semantic alignment, this innovative approach enables us to simultaneously accomplish authentic attribute editing and cross-domain style transfer, a capability not realized in previous methods. Additionally, we ascertain that modifying only the weights of specific layers in the generator can yield an equivalent editing result. Therefore, we introduce an adaptive layer selector, enabling our hypernetworks to autonomously identify the layers requiring output weight factors, which can further improve our hypernetworks' efficiency. Extensive experiments on abundant challenging datasets demonstrate the effectiveness of our method.
Authors: Tao Wu, Tie Luo, Donald C. Wunsch
The capacity to generalize to future unseen data stands as one of the utmost crucial attributes of deep neural networks. Sharpness-Aware Minimization (SAM) aims to enhance the generalizability by minimizing worst-case loss using one-step gradient ascent as an approximation. However, as training progresses, the non-linearity of the loss landscape increases, rendering one-step gradient ascent less effective. On the other hand, multi-step gradient ascent will incur higher training cost. In this paper, we introduce a normalized Hessian trace to accurately measure the curvature of loss landscape on {\em both} training and test sets. In particular, to counter excessive non-linearity of loss landscape, we propose Curvature Regularized SAM (CR-SAM), integrating the normalized Hessian trace as a SAM regularizer. Additionally, we present an efficient way to compute the trace via finite differences with parallelism. Our theoretical analysis based on PAC-Bayes bounds establishes the regularizer's efficacy in reducing generalization error. Empirical evaluation on CIFAR and ImageNet datasets shows that CR-SAM consistently enhances classification performance for ResNet and Vision Transformer (ViT) models across various datasets. Our code is available at https://github.com/TrustAIoT/CR-SAM.
Authors: Pratyusha Sharma, Jordan T. Ash, Dipendra Misra
Transformer-based Large Language Models (LLMs) have become a fixture in modern machine learning. Correspondingly, significant resources are allocated towards research that aims to further advance this technology, typically resulting in models of increasing size that are trained on increasing amounts of data. This work, however, demonstrates the surprising result that it is often possible to significantly improve the performance of LLMs by selectively removing higher-order components of their weight matrices. This simple intervention, which we call LAyer-SElective Rank reduction (LASER), can be done on a model after training has completed, and requires no additional parameters or data. We show extensive experiments demonstrating the generality of this finding across language models and datasets, and provide in-depth analyses offering insights into both when LASER is effective and the mechanism by which it operates.
Authors: Peng Zhao, Jiehua Zhang, Bowen Peng, Longguang Wang, YingMei Wei, Yu Liu, Li Liu
Network binarization exhibits great potential for deployment on resource-constrained devices due to its low computational cost. Despite the critical importance, the security of binarized neural networks (BNNs) is rarely investigated. In this paper, we present ARBiBench, a comprehensive benchmark to evaluate the robustness of BNNs against adversarial perturbations on CIFAR-10 and ImageNet. We first evaluate the robustness of seven influential BNNs on various white-box and black-box attacks. The results reveal that 1) The adversarial robustness of BNNs exhibits a completely opposite performance on the two datasets under white-box attacks. 2) BNNs consistently exhibit better adversarial robustness under black-box attacks. 3) Different BNNs exhibit certain similarities in their robustness performance. Then, we conduct experiments to analyze the adversarial robustness of BNNs based on these insights. Our research contributes to inspiring future research on enhancing the robustness of BNNs and advancing their application in real-world scenarios.
Authors: Chenxu Zhang, Chao Wang, Jianfeng Zhang, Hongyi Xu, Guoxian Song, You Xie, Linjie Luo, Yapeng Tian, Xiaohu Guo, Jiashi Feng
The generation of emotional talking faces from a single portrait image remains a significant challenge. The simultaneous achievement of expressive emotional talking and accurate lip-sync is particularly difficult, as expressiveness is often compromised for the accuracy of lip-sync. As widely adopted by many prior works, the LSTM network often fails to capture the subtleties and variations of emotional expressions. To address these challenges, we introduce DREAM-Talk, a two-stage diffusion-based audio-driven framework, tailored for generating diverse expressions and accurate lip-sync concurrently. In the first stage, we propose EmoDiff, a novel diffusion module that generates diverse highly dynamic emotional expressions and head poses in accordance with the audio and the referenced emotion style. Given the strong correlation between lip motion and audio, we then refine the dynamics with enhanced lip-sync accuracy using audio features and emotion style. To this end, we deploy a video-to-video rendering module to transfer the expressions and lip motions from our proxy 3D avatar to an arbitrary portrait. Both quantitatively and qualitatively, DREAM-Talk outperforms state-of-the-art methods in terms of expressiveness, lip-sync accuracy and perceptual quality.
Authors: Chengen Lai, Shengli Song, Shiqi Meng, Jingyang Li, Sitong Yan, Guangneng Hu
Natural language explanation in visual question answer (VQA-NLE) aims to explain the decision-making process of models by generating natural language sentences to increase users' trust in the black-box systems. Existing post-hoc methods have achieved significant progress in obtaining a plausible explanation. However, such post-hoc explanations are not always aligned with human logical inference, suffering from the issues on: 1) Deductive unsatisfiability, the generated explanations do not logically lead to the answer; 2) Factual inconsistency, the model falsifies its counterfactual explanation for answers without considering the facts in images; and 3) Semantic perturbation insensitivity, the model can not recognize the semantic changes caused by small perturbations. These problems reduce the faithfulness of explanations generated by models. To address the above issues, we propose a novel self-supervised \textbf{M}ulti-level \textbf{C}ontrastive \textbf{L}earning based natural language \textbf{E}xplanation model (MCLE) for VQA with semantic-level, image-level, and instance-level factual and counterfactual samples. MCLE extracts discriminative features and aligns the feature spaces from explanations with visual question and answer to generate more consistent explanations. We conduct extensive experiments, ablation analysis, and case study to demonstrate the effectiveness of our method on two VQA-NLE benchmarks.
Authors: Keqiang Sun, Dor Litvak, Yunzhi Zhang, Hongsheng Li, Jiajun Wu, Shangzhe Wu
We introduce Ponymation, a new method for learning a generative model of articulated 3D animal motions from raw, unlabeled online videos. Unlike existing approaches for motion synthesis, our model does not require any pose annotations or parametric shape models for training, and is learned purely from a collection of raw video clips obtained from the Internet. We build upon a recent work, MagicPony, which learns articulated 3D animal shapes purely from single image collections, and extend it on two fronts. First, instead of training on static images, we augment the framework with a video training pipeline that incorporates temporal regularizations, achieving more accurate and temporally consistent reconstructions. Second, we learn a generative model of the underlying articulated 3D motion sequences via a spatio-temporal transformer VAE, simply using 2D reconstruction losses without relying on any explicit pose annotations. At inference time, given a single 2D image of a new animal instance, our model reconstructs an articulated, textured 3D mesh, and generates plausible 3D animations by sampling from the learned motion latent space.
Authors: Lvyang Yang, Jiankang Zhang, Huaiqiang Li, Longfei Ren, Chen Yang, Jingyu Wang, Dongyuan Shi
The digitization of engineering drawings is crucial for efficient reuse, distribution, and archiving. Existing computer vision approaches for digitizing engineering drawings typically assume the input drawings have high quality. However, in reality, engineering drawings are often blurred and distorted due to improper scanning, storage, and transmission, which may jeopardize the effectiveness of existing approaches. This paper focuses on restoring and recognizing low-quality engineering drawings, where an end-to-end framework is proposed to improve the quality of the drawings and identify the graphical symbols on them. The framework uses K-means clustering to classify different engineering drawing patches into simple and complex texture patches based on their gray level co-occurrence matrix statistics. Computer vision operations and a modified Enhanced Super-Resolution Generative Adversarial Network (ESRGAN) model are then used to improve the quality of the two types of patches, respectively. A modified Faster Region-based Convolutional Neural Network (Faster R-CNN) model is used to recognize the quality-enhanced graphical symbols. Additionally, a multi-stage task-driven collaborative learning strategy is proposed to train the modified ESRGAN and Faster R-CNN models to improve the resolution of engineering drawings in the direction that facilitates graphical symbol recognition, rather than human visual perception. A synthetic data generation method is also proposed to construct quality-degraded samples for training the framework. Experiments on real-world electrical diagrams show that the proposed framework achieves an accuracy of 98.98% and a recall of 99.33%, demonstrating its superiority over previous approaches. Moreover, the framework is integrated into a widely-used power system software application to showcase its practicality.
Authors: Zhiyu Zhu, Huaming Chen, Jiayu Zhang, Xinyi Wang, Zhibo Jin, Minhui Xue, Dongxiao Zhu, Kim-Kwang Raymond Choo
To better understand the output of deep neural networks (DNN), attribution based methods have been an important approach for model interpretability, which assign a score for each input dimension to indicate its importance towards the model outcome. Notably, the attribution methods use the axioms of sensitivity and implementation invariance to ensure the validity and reliability of attribution results. Yet, the existing attribution methods present challenges for effective interpretation and efficient computation. In this work, we introduce MFABA, an attribution algorithm that adheres to axioms, as a novel method for interpreting DNN. Additionally, we provide the theoretical proof and in-depth analysis for MFABA algorithm, and conduct a large scale experiment. The results demonstrate its superiority by achieving over 101.5142 times faster speed than the state-of-the-art attribution algorithms. The effectiveness of MFABA is thoroughly evaluated through the statistical analysis in comparison to other methods, and the full implementation package is open-source at: https://github.com/LMBTough/MFABA
Authors: Jing Li, Qiu-Feng Wang, Kaizhu Huang, Rui Zhang, Siyuan Wang
Deciphering the oracle bone script plays a significant role in Chinese archaeology and philology. However, it is significantly challenging due to the scarcity of oracle character images. To overcome this issue, we propose Diff-Oracle, based on diffusion models (DMs), to generate sufficient controllable oracle characters. In contrast to most DMs that rely on text prompts, we incorporate a style encoder to control style information during the generation process. This encoder extracts style prompts from existing oracle character images, where style details are converted from a CLIP model into a text embedding format. Inspired by ControlNet, we introduce a content encoder to capture desired content information from content images, ensuring the fidelity of character glyphs. To train Diff-Oracle effectively, we propose to obtain pixel-level paired oracle character images (i.e., style and content images) by a pre-trained image-to-image translation model. Extensive qualitative and quantitative experiments conducted on two benchmark datasets, Oracle-241 and OBC306, demonstrate that our Diff-Oracle outperforms existing generative methods in terms of image generation, further enhancing recognition accuracy. Source codes will be available.
Authors: Waris Gill (1), Ali Anwar (2), Muhammad Ali Gulzar (1) ((1) Virginia Tech, (2) University of Minnesota Twin Cities)
Federated Learning (FL) trains a collaborative machine learning model by aggregating multiple privately trained clients' models over several training rounds. Such a long, continuous action of model aggregations poses significant challenges in reasoning about the origin and composition of such a global model. Regardless of the quality of the global model or if it has a fault, understanding the model's origin is equally important for debugging, interpretability, and explainability in federated learning. FL application developers often question: (1) what clients contributed towards a global model and (2) if a global model predicts a label, which clients are responsible for it?
We introduce, neuron provenance, a fine-grained lineage capturing mechanism that tracks the flow of information between the individual participating clients in FL and the final global model. We operationalize this concept in ProvFL that functions on two key principles. First, recognizing that monitoring every neuron of every client's model statically is ineffective and noisy due to the uninterpretable nature of individual neurons, ProvFL dynamically isolates influential and sensitive neurons in the global model, significantly reducing the search space. Second, as multiple clients' models are fused in each round to form a global model, tracking each client's contribution becomes challenging. ProvFL leverages the invertible nature of fusion algorithms to precisely isolate each client's contribution derived from selected neurons. When asked to localize the clients responsible for the given behavior (i.e., prediction) of the global model, ProvFL successfully localizes them with an average provenance accuracy of 97%. Additionally, ProvFL outperforms the state-of-the-art FL fault localization approach by an average margin of 50%.
Authors: Haifeng Huang, Yang Zhao, Zehan Wang, Yan Xia, Zhou Zhao
Temporal Video Grounding (TVG) aims to localize the temporal boundary of a specific segment in an untrimmed video based on a given language query. Since datasets in this domain are often gathered from limited video scenes, models tend to overfit to scene-specific factors, which leads to suboptimal performance when encountering new scenes in real-world applications. In a new scene, the fine-grained annotations are often insufficient due to the expensive labor cost, while the coarse-grained video-query pairs are easier to obtain. Thus, to address this issue and enhance model performance on new scenes, we explore the TVG task in an unsupervised domain adaptation (UDA) setting across scenes for the first time, where the video-query pairs in the source scene (domain) are labeled with temporal boundaries, while those in the target scene are not. Under the UDA setting, we introduce a novel Adversarial Multi-modal Domain Adaptation (AMDA) method to adaptively adjust the model's scene-related knowledge by incorporating insights from the target data. Specifically, we tackle the domain gap by utilizing domain discriminators, which help identify valuable scene-related features effective across both domains. Concurrently, we mitigate the semantic gap between different modalities by aligning video-query pairs with related semantics. Furthermore, we employ a mask-reconstruction approach to enhance the understanding of temporal semantics within a scene. Extensive experiments on Charades-STA, ActivityNet Captions, and YouCook2 demonstrate the effectiveness of our proposed method.
Authors: Yun Zhu, Le Hui, Yaqi Shen, Jin Xie
Current 3D object detection methods for indoor scenes mainly follow the voting-and-grouping strategy to generate proposals. However, most methods utilize instance-agnostic groupings, such as ball query, leading to inconsistent semantic information and inaccurate regression of the proposals. To this end, we propose a novel superpoint grouping network for indoor anchor-free one-stage 3D object detection. Specifically, we first adopt an unsupervised manner to partition raw point clouds into superpoints, areas with semantic consistency and spatial similarity. Then, we design a geometry-aware voting module that adapts to the centerness in anchor-free detection by constraining the spatial relationship between superpoints and object centers. Next, we present a superpoint-based grouping module to explore the consistent representation within proposals. This module includes a superpoint attention layer to learn feature interaction between neighboring superpoints, and a superpoint-voxel fusion layer to propagate the superpoint-level information to the voxel level. Finally, we employ effective multiple matching to capitalize on the dynamic receptive fields of proposals based on superpoints during the training. Experimental results demonstrate our method achieves state-of-the-art performance on ScanNet V2, SUN RGB-D, and S3DIS datasets in the indoor one-stage 3D object detection. Source code is available at https://github.com/zyrant/SPGroup3D.
Authors: Dongseob Kim, Seungho Lee, Junsuk Choe, Hyunjung Shim
State-of-the-art techniques in weakly-supervised semantic segmentation (WSSS) using image-level labels exhibit severe performance degradation on driving scene datasets such as Cityscapes. To address this challenge, we develop a new WSSS framework tailored to driving scene datasets. Based on extensive analysis of dataset characteristics, we employ Contrastive Language-Image Pre-training (CLIP) as our baseline to obtain pseudo-masks. However, CLIP introduces two key challenges: (1) pseudo-masks from CLIP lack in representing small object classes, and (2) these masks contain notable noise. We propose solutions for each issue as follows. (1) We devise Global-Local View Training that seamlessly incorporates small-scale patches during model training, thereby enhancing the model's capability to handle small-sized yet critical objects in driving scenes (e.g., traffic light). (2) We introduce Consistency-Aware Region Balancing (CARB), a novel technique that discerns reliable and noisy regions through evaluating the consistency between CLIP masks and segmentation predictions. It prioritizes reliable pixels over noisy pixels via adaptive loss weighting. Notably, the proposed method achieves 51.8\% mIoU on the Cityscapes test dataset, showcasing its potential as a strong WSSS baseline on driving scene datasets. Experimental results on CamVid and WildDash2 demonstrate the effectiveness of our method across diverse datasets, even with small-scale datasets or visually challenging conditions. The code is available at https://github.com/k0u-id/CARB.
Authors: Peng Gao (1), Ahmed Jaafar (1), Brian Reily (2), Christopher Reardon (3), Hao Zhang (1) ((1) University of Massachusetts Amherst, (2) DEVCOM Army Research Laboratory, (3) University of Denver)
Language-enabled robots have been widely studied over the past years to enable natural human-robot interaction and teaming in various real-world applications. Language-enabled robots must be able to comprehend referring expressions to identify a particular object from visual perception using a set of referring attributes extracted from natural language. However, visual observations of an object may not be available when it is referred to, and the number of objects and attributes may also be unbounded in open worlds. To address the challenges, we implement an attribute-based compositional zero-shot learning method that uses a list of attributes to perform referring expression comprehension in open worlds. We evaluate the approach on two datasets including the MIT-States and the Clothing 16K. The preliminary experimental results show that our implemented approach allows a robot to correctly identify the objects referred to by human commands.
Authors: Nazmul Karim, Umar Khalid, Hasan Iqbal, Jing Hua, Chen Chen
Text-to-Image (T2I) diffusion models have gained popularity recently due to their multipurpose and easy-to-use nature, e.g. image and video generation as well as editing. However, training a diffusion model specifically for 3D scene editing is not straightforward due to the lack of large-scale datasets. To date, editing 3D scenes requires either re-training the model to adapt to various 3D edited scenes or design-specific methods for each special editing type. Furthermore, state-of-the-art (SOTA) methods require multiple synchronized edited images from the same scene to facilitate the scene editing. Due to the current limitations of T2I models, it is very challenging to apply consistent editing effects to multiple images, i.e. multi-view inconsistency in editing. This in turn compromises the desired 3D scene editing performance if these images are used. In our work, we propose a novel training-free 3D scene editing technique, Free-Editor, which allows users to edit 3D scenes without further re-training the model during test time. Our proposed method successfully avoids the multi-view style inconsistency issue in SOTA methods with the help of a "single-view editing" scheme. Specifically, we show that editing a particular 3D scene can be performed by only modifying a single view. To this end, we introduce an Edit Transformer that enforces intra-view consistency and inter-view style transfer by utilizing self- and cross-attention, respectively. Since it is no longer required to re-train the model and edit every view in a scene, the editing time, as well as memory resources, are reduced significantly, e.g., the runtime being $\sim \textbf{20} \times$ faster than SOTA. We have conducted extensive experiments on a wide range of benchmark datasets and achieve diverse editing capabilities with our proposed technique.
Authors: Miao Hua, Jiawei Liu, Fei Ding, Wei Liu, Jie Wu, Qian He
Diffusion-based models have demonstrated impressive capabilities for text-to-image generation and are expected for personalized applications of subject-driven generation, which require the generation of customized concepts with one or a few reference images. However, existing methods based on fine-tuning fail to balance the trade-off between subject learning and the maintenance of the generation capabilities of pretrained models. Moreover, other methods that utilize additional image encoders tend to lose important details of the subject due to encoding compression. To address these challenges, we propose DreamTurner, a novel method that injects reference information from coarse to fine to achieve subject-driven image generation more effectively. DreamTurner introduces a subject-encoder for coarse subject identity preservation, where the compressed general subject features are introduced through an attention layer before visual-text cross-attention. We then modify the self-attention layers within pretrained text-to-image models to self-subject-attention layers to refine the details of the target subject. The generated image queries detailed features from both the reference image and itself in self-subject-attention. It is worth emphasizing that self-subject-attention is an effective, elegant, and training-free method for maintaining the detailed features of customized subjects and can serve as a plug-and-play solution during inference. Finally, with additional subject-driven fine-tuning, DreamTurner achieves remarkable performance in subject-driven image generation, which can be controlled by a text or other conditions such as pose. For further details, please visit the project page at https://dreamtuner-diffusion.github.io/.
Authors: Haochen Wang, Junsong Fan, Yuxi Wang, Kaiyou Song, Tiancai Wang, Xiangyu Zhang, Zhaoxiang Zhang
Masked visual modeling has attracted much attention due to its promising potential in learning generalizable representations. Typical approaches urge models to predict specific contents of masked tokens, which can be intuitively considered as teaching a student (the model) to solve given problems (predicting masked contents). Under such settings, the performance is highly correlated with mask strategies (the difficulty of provided problems). We argue that it is equally important for the model to stand in the shoes of a teacher to produce challenging problems by itself. Intuitively, patches with high values of reconstruction loss can be regarded as hard samples, and masking those hard patches naturally becomes a demanding reconstruction task. To empower the model as a teacher, we propose Hard Patches Mining (HPM), predicting patch-wise losses and subsequently determining where to mask. Technically, we introduce an auxiliary loss predictor, which is trained with a relative objective to prevent overfitting to exact loss values. Also, to gradually guide the training procedure, we propose an easy-to-hard mask strategy. Empirically, HPM brings significant improvements under both image and video benchmarks. Interestingly, solely incorporating the extra loss prediction objective leads to better representations, verifying the efficacy of determining where is hard to reconstruct. The code is available at https://github.com/Haochen-Wang409/HPM.
Authors: Dawid Malarz, Weronika Smolak, Jacek Tabor, Sławomir Tadeja, Przemysław Spurek
Neural Radiance Fields (NeRFs) have demonstrated the remarkable potential of neural networks to capture the intricacies of 3D objects. By encoding the shape and color information within neural network weights, NeRFs excel at producing strikingly sharp novel views of 3D objects. Recently, numerous generalizations of NeRFs utilizing generative models have emerged, expanding its versatility. In contrast, Gaussian Splatting (GS) offers a similar renders quality with faster training and inference as it does not need neural networks to work. We encode information about the 3D objects in the set of Gaussian distributions that can be rendered in 3D similarly to classical meshes. Unfortunately, GS are difficult to condition since they usually require circa hundred thousand Gaussian components. To mitigate the caveats of both models, we propose a hybrid model that uses GS representation of the 3D object's shape and NeRF-based encoding of color and opacity. Our model uses Gaussian distributions with trainable positions (i.e. means of Gaussian), shape (i.e. covariance of Gaussian), color and opacity, and neural network, which takes parameters of Gaussian and viewing direction to produce changes in color and opacity. Consequently, our model better describes shadows, light reflections, and transparency of 3D objects.
Authors: Xinghao Chen, Siwei Li, Yijing Yang, Yunhe Wang
Detection Transformer (DETR) and its variants have shown great potential for accurate object detection in recent years. The mechanism of object query enables DETR family to directly obtain a fixed number of object predictions and streamlines the detection pipeline. Meanwhile, recent studies also reveal that with proper architecture design, convolution networks (ConvNets) also achieve competitive performance with transformers, \eg, ConvNeXt. To this end, in this paper we explore whether we could build a query-based end-to-end object detection framework with ConvNets instead of sophisticated transformer architecture. The proposed framework, \ie, Detection ConvNet (DECO), is composed of a backbone and convolutional encoder-decoder architecture. We carefully design the DECO encoder and propose a novel mechanism for our DECO decoder to perform interaction between object queries and image features via convolutional layers. We compare the proposed DECO against prior detectors on the challenging COCO benchmark. Despite its simplicity, our DECO achieves competitive performance in terms of detection accuracy and running speed. Specifically, with the ResNet-50 and ConvNeXt-Tiny backbone, DECO obtains $38.6\%$ and $40.8\%$ AP on COCO \textit{val} set with $35$ and $28$ FPS respectively and outperforms the DETR model. Incorporated with advanced multi-scale feature module, our DECO+ achieves $47.8\%$ AP with $34$ FPS. We hope the proposed DECO brings another perspective for designing object detection framework.
Authors: Mingfei Han, Linjie Yang, Xiaojie Jin, Jiashi Feng, Xiaojun Chang, Heng Wang
The creation of new datasets often presents new challenges for video recognition and can inspire novel ideas while addressing these challenges. While existing datasets mainly comprise landscape mode videos, our paper seeks to introduce portrait mode videos to the research community and highlight the unique challenges associated with this video format. With the growing popularity of smartphones and social media applications, recognizing portrait mode videos is becoming increasingly important. To this end, we have developed the first dataset dedicated to portrait mode video recognition, namely PortraitMode-400. The taxonomy of PortraitMode-400 was constructed in a data-driven manner, comprising 400 fine-grained categories, and rigorous quality assurance was implemented to ensure the accuracy of human annotations. In addition to the new dataset, we conducted a comprehensive analysis of the impact of video format (portrait mode versus landscape mode) on recognition accuracy and spatial bias due to the different formats. Furthermore, we designed extensive experiments to explore key aspects of portrait mode video recognition, including the choice of data augmentation, evaluation procedure, the importance of temporal information, and the role of audio modality. Building on the insights from our experimental results and the introduction of PortraitMode-400, our paper aims to inspire further research efforts in this emerging research area.
Authors: Yang Nan, Xiaodan Xing, Shiyi Wang, Zeyu Tang, Federico N Felder, Sheng Zhang, Roberta Eufrasia Ledda, Xiaoliu Ding, Ruiqi Yu, Weiping Liu, Feng Shi, Tianyang Sun, Zehong Cao, Minghui Zhang, Yun Gu, Hanxiao Zhang, Jian Gao, Wen Tang, Pengxin Yu, Han Kang, Junqiang Chen, Xing Lu, Boyu Zhang, Michail Mamalakis, Francesco Prinzi, Gianluca Carlini, Lisa Cuneo, Abhirup Banerjee, Zhaohu Xing, Lei Zhu, Zacharia Mesbah, Dhruv Jain, Tsiry Mayet, Hongyu Yuan, Qing Lyu, Athol Wells, Simon LF Walsh, Guang Yang
Airway-related quantitative imaging biomarkers are crucial for examination, diagnosis, and prognosis in pulmonary diseases. However, the manual delineation of airway trees remains prohibitively time-consuming. While significant efforts have been made towards enhancing airway modelling, current public-available datasets concentrate on lung diseases with moderate morphological variations. The intricate honeycombing patterns present in the lung tissues of fibrotic lung disease patients exacerbate the challenges, often leading to various prediction errors. To address this issue, the 'Airway-Informed Quantitative CT Imaging Biomarker for Fibrotic Lung Disease 2023' (AIIB23) competition was organized in conjunction with the official 2023 International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI). The airway structures were meticulously annotated by three experienced radiologists. Competitors were encouraged to develop automatic airway segmentation models with high robustness and generalization abilities, followed by exploring the most correlated QIB of mortality prediction. A training set of 120 high-resolution computerised tomography (HRCT) scans were publicly released with expert annotations and mortality status. The online validation set incorporated 52 HRCT scans from patients with fibrotic lung disease and the offline test set included 140 cases from fibrosis and COVID-19 patients. The results have shown that the capacity of extracting airway trees from patients with fibrotic lung disease could be enhanced by introducing voxel-wise weighted general union loss and continuity loss. In addition to the competitive image biomarkers for prognosis, a strong airway-derived biomarker (Hazard ratio>1.5, p<0.0001) was revealed for survival prognostication compared with existing clinical measurements, clinician assessment and AI-based biomarkers.
Authors: Huan Ling, Seung Wook Kim, Antonio Torralba, Sanja Fidler, Karsten Kreis
Text-guided diffusion models have revolutionized image and video generation and have also been successfully used for optimization-based 3D object synthesis. Here, we instead focus on the underexplored text-to-4D setting and synthesize dynamic, animated 3D objects using score distillation methods with an additional temporal dimension. Compared to previous work, we pursue a novel compositional generation-based approach, and combine text-to-image, text-to-video, and 3D-aware multiview diffusion models to provide feedback during 4D object optimization, thereby simultaneously enforcing temporal consistency, high-quality visual appearance and realistic geometry. Our method, called Align Your Gaussians (AYG), leverages dynamic 3D Gaussian Splatting with deformation fields as 4D representation. Crucial to AYG is a novel method to regularize the distribution of the moving 3D Gaussians and thereby stabilize the optimization and induce motion. We also propose a motion amplification mechanism as well as a new autoregressive synthesis scheme to generate and combine multiple 4D sequences for longer generation. These techniques allow us to synthesize vivid dynamic scenes, outperform previous work qualitatively and quantitatively and achieve state-of-the-art text-to-4D performance. Due to the Gaussian 4D representation, different 4D animations can be seamlessly combined, as we demonstrate. AYG opens up promising avenues for animation, simulation and digital content creation as well as synthetic data generation.
Authors: Junfei Xiao, Ziqi Zhou, Wenxuan Li, Shiyi Lan, Jieru Mei, Zhiding Yu, Alan Yuille, Yuyin Zhou, Cihang Xie
This paper introduces ProLab, a novel approach using property-level label space for creating strong interpretable segmentation models. Instead of relying solely on category-specific annotations, ProLab uses descriptive properties grounded in common sense knowledge for supervising segmentation models. It is based on two core designs. First, we employ Large Language Models (LLMs) and carefully crafted prompts to generate descriptions of all involved categories that carry meaningful common sense knowledge and follow a structured format. Second, we introduce a description embedding model preserving semantic correlation across descriptions and then cluster them into a set of descriptive properties (e.g., 256) using K-Means. These properties are based on interpretable common sense knowledge consistent with theories of human recognition. We empirically show that our approach makes segmentation models perform stronger on five classic benchmarks (e.g., ADE20K, COCO-Stuff, Pascal Context, Cityscapes, and BDD). Our method also shows better scalability with extended training steps than category-level supervision. Our interpretable segmentation framework also emerges with the generalization ability to segment out-of-domain or unknown categories using only in-domain descriptive properties. Code is available at https://github.com/lambert-x/ProLab.
Authors: Zheheng Jiang, Hossein Rahmani, Sue Black, Bryan M. Williams
We present 3D Points Splatting Hand Reconstruction (3D-PSHR), a real-time and photo-realistic hand reconstruction approach. We propose a self-adaptive canonical points upsampling strategy to achieve high-resolution hand geometry representation. This is followed by a self-adaptive deformation that deforms the hand from the canonical space to the target pose, adapting to the dynamic changing of canonical points which, in contrast to the common practice of subdividing the MANO model, offers greater flexibility and results in improved geometry fitting. To model texture, we disentangle the appearance color into the intrinsic albedo and pose-aware shading, which are learned through a Context-Attention module. Moreover, our approach allows the geometric and the appearance models to be trained simultaneously in an end-to-end manner. We demonstrate that our method is capable of producing animatable, photorealistic and relightable hand reconstructions using multiple datasets, including monocular videos captured with handheld smartphones and large-scale multi-view videos featuring various hand poses. We also demonstrate that our approach achieves real-time rendering speeds while simultaneously maintaining superior performance compared to existing state-of-the-art methods.
Authors: Zhao Yang, Jiaxuan Liu, Yucheng Han, Xin Chen, Zebiao Huang, Bin Fu, Gang Yu
Recent advancements in large language models (LLMs) have led to the creation of intelligent agents capable of performing complex tasks. This paper introduces a novel LLM-based multimodal agent framework designed to operate smartphone applications. Our framework enables the agent to operate smartphone applications through a simplified action space, mimicking human-like interactions such as tapping and swiping. This novel approach bypasses the need for system back-end access, thereby broadening its applicability across diverse apps. Central to our agent's functionality is its innovative learning method. The agent learns to navigate and use new apps either through autonomous exploration or by observing human demonstrations. This process generates a knowledge base that the agent refers to for executing complex tasks across different applications. To demonstrate the practicality of our agent, we conducted extensive testing over 50 tasks in 10 different applications, including social media, email, maps, shopping, and sophisticated image editing tools. The results affirm our agent's proficiency in handling a diverse array of high-level tasks.
Authors: Haozheng Zhang, Edmond S. L. Ho, Xiatian Zhang, Silvia Del Din, Hubert P. H. Shum
Purpose:Current methods for diagnosis of PD rely on clinical examination. The accuracy of diagnosis ranges between 73% and 84%, and is influenced by the experience of the clinical assessor. Hence, an automatic, effective and interpretable supporting system for PD symptom identification would support clinicians in making more robust PD diagnostic decisions. Methods: We propose to analyze Parkinson's tremor (PT) to support the analysis of PD, since PT is one of the most typical symptoms of PD with broad generalizability. To realize the idea, we present SPA-PTA, a deep learning-based PT classification and severity estimation system that takes consumer-grade videos of front-facing humans as input. The core of the system is a novel attention module with a lightweight pyramidal channel-squeezing-fusion architecture that effectively extracts relevant PT information and filters noise. It enhances modeling performance while improving system interpretability. Results:We validate our system via individual-based leave-one-out cross-validation on two tasks: the PT classification task and the tremor severity rating estimation task. Our system presents a 91.3% accuracy and 80.0% F1-score in classifying PT with non-PT class, while providing a 76.4% accuracy and 76.7% F1-score in more complex multiclass tremor rating classification task. Conclusion: Our system offers a cost-effective PT classification and tremor severity estimation results as warning signs of PD for undiagnosed patients with PT symptoms. In addition, it provides a potential solution for supporting PD diagnosis in regions with limited clinical resources.
Authors: Linger Deng, Mingxin Huang, Xudong Xie, Yuliang Liu, Lianwen Jin, Xiang Bai
The advancement of text shape representations towards compactness has enhanced text detection and spotting performance, but at a high annotation cost. Current models use single-point annotations to reduce costs, yet they lack sufficient localization information for downstream applications. To overcome this limitation, we introduce Point2Polygon, which can efficiently transform single-points into compact polygons. Our method uses a coarse-to-fine process, starting with creating and selecting anchor points based on recognition confidence, then vertically and horizontally refining the polygon using recognition information to optimize its shape. We demonstrate the accuracy of the generated polygons through extensive experiments: 1) By creating polygons from ground truth points, we achieved an accuracy of 82.0% on ICDAR 2015; 2) In training detectors with polygons generated by our method, we attained 86% of the accuracy relative to training with ground truth (GT); 3) Additionally, the proposed Point2Polygon can be seamlessly integrated to empower single-point spotters to generate polygons. This integration led to an impressive 82.5% accuracy for the generated polygons. It is worth mentioning that our method relies solely on synthetic recognition information, eliminating the need for any manual annotation beyond single points.
Authors: Soopil Kim, Sion An, Philip Chikontwe, Myeongkyun Kang, Ehsan Adeli, Kilian M. Pohl, Sanghyun Park
Logical anomalies (LA) refer to data violating underlying logical constraints e.g., the quantity, arrangement, or composition of components within an image. Detecting accurately such anomalies requires models to reason about various component types through segmentation. However, curation of pixel-level annotations for semantic segmentation is both time-consuming and expensive. Although there are some prior few-shot or unsupervised co-part segmentation algorithms, they often fail on images with industrial object. These images have components with similar textures and shapes, and a precise differentiation proves challenging. In this study, we introduce a novel component segmentation model for LA detection that leverages a few labeled samples and unlabeled images sharing logical constraints. To ensure consistent segmentation across unlabeled images, we employ a histogram matching loss in conjunction with an entropy loss. As segmentation predictions play a crucial role, we propose to enhance both local and global sample validity detection by capturing key aspects from visual semantics via three memory banks: class histograms, component composition embeddings and patch-level representations. For effective LA detection, we propose an adaptive scaling strategy to standardize anomaly scores from different memory banks in inference. Extensive experiments on the public benchmark MVTec LOCO AD reveal our method achieves 98.1% AUROC in LA detection vs. 89.6% from competing methods.
Authors: Han Shu, Wenshuo Li, Yehui Tang, Yiman Zhang, Yihao Chen, Houqiang Li, Yunhe Wang, Xinghao Chen
Recently segment anything model (SAM) has shown powerful segmentation capability and has drawn great attention in computer vision fields. Massive following works have developed various applications based on the pretrained SAM and achieved impressive performance on downstream vision tasks. However, SAM consists of heavy architectures and requires massive computational capacity, which hinders the further application of SAM on computation constrained edge devices. To this end, in this paper we propose a framework to obtain a tiny segment anything model (TinySAM) while maintaining the strong zero-shot performance. We first propose a full-stage knowledge distillation method with online hard prompt sampling strategy to distill a lightweight student model. We also adapt the post-training quantization to the promptable segmentation task and further reduce the computational cost. Moreover, a hierarchical segmenting everything strategy is proposed to accelerate the everything inference by $2\times$ with almost no performance degradation. With all these proposed methods, our TinySAM leads to orders of magnitude computational reduction and pushes the envelope for efficient segment anything task. Extensive experiments on various zero-shot transfer tasks demonstrate the significantly advantageous performance of our TinySAM against counterpart methods. Pre-trained models and codes will be available at https://github.com/xinghaochen/TinySAM and https://gitee.com/mindspore/models/tree/master/research/cv/TinySAM.
Authors: Marvin Kahra, Michael Breuß, Andreas Kleefeld, Martin Welk
Mathematical morphology is a part of image processing that has proven to be fruitful for numerous applications. Two main operations in mathematical morphology are dilation and erosion. These are based on the construction of a supremum or infimum with respect to an order over the tonal range in a certain section of the image. The tonal ordering can easily be realised in grey-scale morphology, and some morphological methods have been proposed for colour morphology. However, all of these have certain limitations. In this paper we present a novel approach to colour morphology extending upon previous work in the field based on the Loewner order. We propose to consider an approximation of the supremum by means of a log-sum exponentiation introduced by Maslov. We apply this to the embedding of an RGB image in a field of symmetric $2\times2$ matrices. In this way we obtain nearly isotropic matrices representing colours and the structural advantage of transitivity. In numerical experiments we highlight some remarkable properties of the proposed approach.
Authors: Marina Ljubenovic, Alessia Artesani, Stefano Bonetti, Arianna Traviglia
Terahertz time-domain spectroscopy (THz-TDS) employs sub-picosecond pulses to probe dielectric properties of materials giving as a result a 3-dimensional hyperspectral data cube. The spatial resolution of THz images is primarily limited by two sources: a non-zero THz beam waist and the acquisition step size. Acquisition with a small step size allows for the visualisation of smaller details in images at the expense of acquisition time, but the frequency-dependent point-spread function remains the biggest bottleneck for THz imaging. This work presents a super-resolution approach to restore THz time-domain images acquired with medium-to-big step sizes. The results show the optimized and robust performance for different frequency bands (from 0.5 to 3.5 THz) obtaining higher resolution and additionally removing effects of blur at lower frequencies and noise at higher frequencies.
Authors: Kwangrok Ryoo, Yeonsik Jo, Seungjun Lee, Mira Kim, Ahra Jo, Seung Hwan Kim, Seungryong Kim, Soonyoung Lee
For object detection task with noisy labels, it is important to consider not only categorization noise, as in image classification, but also localization noise, missing annotations, and bogus bounding boxes. However, previous studies have only addressed certain types of noise (e.g., localization or categorization). In this paper, we propose Universal-Noise Annotation (UNA), a more practical setting that encompasses all types of noise that can occur in object detection, and analyze how UNA affects the performance of the detector. We analyzed the development direction of previous works of detection algorithms and examined the factors that impact the robustness of detection model learning method. We open-source the code for injecting UNA into the dataset and all the training log and weight are also shared.
Authors: Ahmet Haydar Ornek, Deniz Sen, Esmanur Civil
The main aim of this study is to demonstrate how innovative view synthesis and 3D reconstruction techniques can be used to create models of endangered species using monocular RGB images. To achieve this, we employed SyncDreamer to produce unique perspectives and NeuS and NeRF to reconstruct 3D representations. We chose four different animals, including the oriental stork, frog, dragonfly, and tiger, as our subjects for this study. Our results show that the combination of SyncDreamer, NeRF, and NeuS techniques can successfully create 3D models of endangered animals. However, we also observed that NeuS produced blurry images, while NeRF generated sharper but noisier images. This study highlights the potential of modeling endangered animals and offers a new direction for future research in this field. By showcasing the effectiveness of these advanced techniques, we hope to encourage further exploration and development of techniques for preserving and studying endangered species.
Authors: Bichen Wu, Ching-Yao Chuang, Xiaoyan Wang, Yichen Jia, Kapil Krishnakumar, Tong Xiao, Feng Liang, Licheng Yu, Peter Vajda
In this paper, we introduce Fairy, a minimalist yet robust adaptation of image-editing diffusion models, enhancing them for video editing applications. Our approach centers on the concept of anchor-based cross-frame attention, a mechanism that implicitly propagates diffusion features across frames, ensuring superior temporal coherence and high-fidelity synthesis. Fairy not only addresses limitations of previous models, including memory and processing speed. It also improves temporal consistency through a unique data augmentation strategy. This strategy renders the model equivariant to affine transformations in both source and target images. Remarkably efficient, Fairy generates 120-frame 512x384 videos (4-second duration at 30 FPS) in just 14 seconds, outpacing prior works by at least 44x. A comprehensive user study, involving 1000 generated samples, confirms that our approach delivers superior quality, decisively outperforming established methods.
Authors: Thomas Norrenbrock, Marco Rudolph, Bodo Rosenhahn
Explanations in Computer Vision are often desired, but most Deep Neural Networks can only provide saliency maps with questionable faithfulness. Self-Explaining Neural Networks (SENN) extract interpretable concepts with fidelity, diversity, and grounding to combine them linearly for decision-making. While they can explain what was recognized, initial realizations lack accuracy and general applicability. We propose the Quantized-Self-Explaining Neural Network Q-SENN. Q-SENN satisfies or exceeds the desiderata of SENN while being applicable to more complex datasets and maintaining most or all of the accuracy of an uninterpretable baseline model, out-performing previous work in all considered metrics. Q-SENN describes the relationship between every class and feature as either positive, negative or neutral instead of an arbitrary number of possible relations, enforcing more binary human-friendly features. Since every class is assigned just 5 interpretable features on average, Q-SENN shows convincing local and global interpretability. Additionally, we propose a feature alignment method, capable of aligning learned features with human language-based concepts without additional supervision. Thus, what is learned can be more easily verbalized. The code is published: https://github.com/ThomasNorr/Q-SENN
Authors: Alexander Köhler, Michael Breuß
The computation of correspondences between shapes is a principal task in shape analysis. To this end, methods based on partial differential equations (PDEs) have been established, encompassing e.g. the classic heat kernel signature as well as numerical solution schemes for geometric PDEs. In this work we focus on the latter approach.
We consider here several time stepping schemes. The goal of this investigation is to assess, if one may identify a useful property of methods for time integration for the shape analysis context. Thereby we investigate the dependence on time step size, since the class of implicit schemes that are useful candidates in this context should ideally yield an invariant behaviour with respect to this parameter.
To this end we study integration of heat and wave equation on a manifold. In order to facilitate this study, we propose an efficient, unified model order reduction framework for these models. We show that specific $l_0$ stable schemes are favourable for numerical shape analysis. We give an experimental evaluation of the methods at hand of classical TOSCA data sets.
Authors: Abraham Woubie, Enoch Solomon, Eyael Solomon Emiru
In various verification systems, Restricted Boltzmann Machines (RBMs) have demonstrated their efficacy in both front-end and back-end processes. In this work, we propose the use of RBMs to the image clustering tasks. RBMs are trained to convert images into image embeddings. We employ the conventional bottom-up Agglomerative Hierarchical Clustering (AHC) technique. To address the challenge of limited test face image data, we introduce Agglomerative Hierarchical Clustering based Method for Image Clustering using Restricted Boltzmann Machine (AHC-RBM) with two major steps. Initially, a universal RBM model is trained using all available training dataset. Subsequently, we train an adapted RBM model using the data from each test image. Finally, RBM vectors which is the embedding vector is generated by concatenating the visible-to-hidden weight matrices of these adapted models, and the bias vectors. These vectors effectively preserve class-specific information and are utilized in image clustering tasks. Our experimental results, conducted on two benchmark image datasets (MS-Celeb-1M and DeepFashion), demonstrate that our proposed approach surpasses well-known clustering algorithms such as k-means, spectral clustering, and approximate Rank-order.
Authors: Yimin Sun, Chao Wang, Yan Peng
The zero-shot performance of visual question answering (VQA) models relies heavily on prompts. For example, a zero-shot VQA for disaster scenarios could leverage well-designed Chain of Thought (CoT) prompts to stimulate the model's potential. However, using CoT prompts has some problems, such as causing an incorrect answer in the end due to the hallucination in the thought process. In this paper, we propose a zero-shot VQA named Flood Disaster VQA with Two-Stage Prompt (VQA-TSP). The model generates the thought process in the first stage and then uses the thought process to generate the final answer in the second stage. In particular, visual context is added in the second stage to relieve the hallucination problem that exists in the thought process. Experimental results show that our method exceeds the performance of state-of-the-art zero-shot VQA models for flood disaster scenarios in total. Our study provides a research basis for improving the performance of CoT-based zero-shot VQA.
Authors: Benjamin Alt, Minh Dang Nguyen, Andreas Hermann, Darko Katic, Rainer Jäkel, Rüdiger Dillmann, Eric Sax
The use of autonomous robots for assistance tasks in hospitals has the potential to free up qualified staff and im-prove patient care. However, the ubiquity of deformable and transparent objects in hospital settings poses signif-icant challenges to vision-based perception systems. We present EfficientPPS, a neural architecture for part-aware panoptic segmentation that provides robots with semantically rich visual information for grasping and ma-nipulation tasks. We also present an unsupervised data collection and labelling method to reduce the need for human involvement in the training process. EfficientPPS is evaluated on a dataset containing real-world hospital objects and demonstrated to be robust and efficient in grasping transparent transfusion bags with a collaborative robot arm.
Authors: Xianfang Zeng
This paper presents Paint3D, a novel coarse-to-fine generative framework that is capable of producing high-resolution, lighting-less, and diverse 2K UV texture maps for untextured 3D meshes conditioned on text or image inputs. The key challenge addressed is generating high-quality textures without embedded illumination information, which allows the textures to be re-lighted or re-edited within modern graphics pipelines. To achieve this, our method first leverages a pre-trained depth-aware 2D diffusion model to generate view-conditional images and perform multi-view texture fusion, producing an initial coarse texture map. However, as 2D models cannot fully represent 3D shapes and disable lighting effects, the coarse texture map exhibits incomplete areas and illumination artifacts. To resolve this, we train separate UV Inpainting and UVHD diffusion models specialized for the shape-aware refinement of incomplete areas and the removal of illumination artifacts. Through this coarse-to-fine process, Paint3D can produce high-quality 2K UV textures that maintain semantic consistency while being lighting-less, significantly advancing the state-of-the-art in texturing 3D objects.
Authors: Xiaolong Shen, Jianxin Ma, Chang Zhou, Zongxin Yang
Generating photorealistic 3D faces from given conditions is a challenging task. Existing methods often rely on time-consuming one-by-one optimization approaches, which are not efficient for modeling the same distribution content, e.g., faces. Additionally, an ideal controllable 3D face generation model should consider both facial attributes and expressions. Thus we propose a novel approach called TEx-Face(TExt & Expression-to-Face) that addresses these challenges by dividing the task into three components, i.e., 3D GAN Inversion, Conditional Style Code Diffusion, and 3D Face Decoding. For 3D GAN inversion, we introduce two methods which aim to enhance the representation of style codes and alleviate 3D inconsistencies. Furthermore, we design a style code denoiser to incorporate multiple conditions into the style code and propose a data augmentation strategy to address the issue of insufficient paired visual-language data. Extensive experiments conducted on FFHQ, CelebA-HQ, and CelebA-Dialog demonstrate the promising performance of our TEx-Face in achieving the efficient and controllable generation of photorealistic 3D faces. The code will be available at https://github.com/sxl142/TEx-Face.
Authors: Yiming Zhang, Zhening Xing, Yanhong Zeng, Youqing Fang, Kai Chen
Recent advancements in personalized text-to-image (T2I) models have revolutionized content creation, empowering non-experts to generate stunning images with unique styles. While promising, adding realistic motions into these personalized images by text poses significant challenges in preserving distinct styles, high-fidelity details, and achieving motion controllability by text. In this paper, we present PIA, a Personalized Image Animator that excels in aligning with condition images, achieving motion controllability by text, and the compatibility with various personalized T2I models without specific tuning. To achieve these goals, PIA builds upon a base T2I model with well-trained temporal alignment layers, allowing for the seamless transformation of any personalized T2I model into an image animation model. A key component of PIA is the introduction of the condition module, which utilizes the condition frame and inter-frame affinity as input to transfer appearance information guided by the affinity hint for individual frame synthesis in the latent space. This design mitigates the challenges of appearance-related image alignment within and allows for a stronger focus on aligning with motion-related guidance.
Authors: Han Huang, Yulun Wu, Junsheng Zhou, Ge Gao, Ming Gu, Yushen Liu
Recently, neural implicit functions have demonstrated remarkable results in the field of multi-view reconstruction. However, most existing methods are tailored for dense views and exhibit unsatisfactory performance when dealing with sparse views. Several latest methods have been proposed for generalizing implicit reconstruction to address the sparse view reconstruction task, but they still suffer from high training costs and are merely valid under carefully selected perspectives. In this paper, we propose a novel sparse view reconstruction framework that leverages on-surface priors to achieve highly faithful surface reconstruction. Specifically, we design several constraints on global geometry alignment and local geometry refinement for jointly optimizing coarse shapes and fine details. To achieve this, we train a neural network to learn a global implicit field from the on-surface points obtained from SfM and then leverage it as a coarse geometric constraint. To exploit local geometric consistency, we project on-surface points onto seen and unseen views, treating the consistent loss of projected features as a fine geometric constraint. The experimental results with DTU and BlendedMVS datasets in two prevalent sparse settings demonstrate significant improvements over the state-of-the-art methods.
Authors: Desai Xie, Jiahao Li, Hao Tan, Xin Sun, Zhixin Shu, Yi Zhou, Sai Bi, Sören Pirk, Arie E. Kaufman
Recent advancements in the text-to-3D task leverage finetuned text-to-image diffusion models to generate multi-view images, followed by NeRF reconstruction. Yet, existing supervised finetuned (SFT) diffusion models still suffer from multi-view inconsistency and the resulting NeRF artifacts. Although training longer with SFT improves consistency, it also causes distribution shift, which reduces diversity and realistic details. We argue that the SFT of multi-view diffusion models resembles the instruction finetuning stage of the LLM alignment pipeline and can benefit from RL finetuning (RLFT) methods. Essentially, RLFT methods optimize models beyond their SFT data distribution by using their own outputs, effectively mitigating distribution shift. To this end, we introduce Carve3D, a RLFT method coupled with the Multi-view Reconstruction Consistency (MRC) metric, to improve the consistency of multi-view diffusion models. To compute MRC on a set of multi-view images, we compare them with their corresponding renderings of the reconstructed NeRF at the same viewpoints. We validate the robustness of MRC with extensive experiments conducted under controlled inconsistency levels. We enhance the base RLFT algorithm to stabilize the training process, reduce distribution shift, and identify scaling laws. Through qualitative and quantitative experiments, along with a user study, we demonstrate Carve3D's improved multi-view consistency, the resulting superior NeRF reconstruction quality, and minimal distribution shift compared to longer SFT. Project webpage: https://desaixie.github.io/carve-3d.
Authors: Reuben Markham, Juan M. Espin, Mario Nieto-Hidalgo, Juan E. Tapia
The accurate detection of ID card Presentation Attacks (PA) is becoming increasingly important due to the rising number of online/remote services that require the presentation of digital photographs of ID cards for digital onboarding or authentication. Furthermore, cybercriminals are continuously searching for innovative ways to fool authentication systems to gain unauthorized access to these services. Although advances in neural network design and training have pushed image classification to the state of the art, one of the main challenges faced by the development of fraud detection systems is the curation of representative datasets for training and evaluation. The handcrafted creation of representative presentation attack samples often requires expertise and is very time-consuming, thus an automatic process of obtaining high-quality data is highly desirable. This work explores ID card Presentation Attack Instruments (PAI) in order to improve the generation of samples with four Generative Adversarial Networks (GANs) based image translation models and analyses the effectiveness of the generated data for training fraud detection systems. Using open-source data, we show that synthetic attack presentations are an adequate complement for additional real attack presentations, where we obtain an EER performance increase of 0.63% points for print attacks and a loss of 0.29% for screen capture attacks.
Authors: Enoch Solomon, Abraham Woubie, Eyael Solomon Emiru
Achieving state-of-the-art results in face verification systems typically hinges on the availability of labeled face training data, a resource that often proves challenging to acquire in substantial quantities. In this research endeavor, we proposed employing Siamese networks for face recognition, eliminating the need for labeled face images. We achieve this by strategically leveraging negative samples alongside nearest neighbor counterparts, thereby establishing positive and negative pairs through an unsupervised methodology. The architectural framework adopts a VGG encoder, trained as a double branch siamese network. Our primary aim is to circumvent the necessity for labeled face image data, thus proposing the generation of training pairs in an entirely unsupervised manner. Positive training data are selected within a dataset based on their highest cosine similarity scores with a designated anchor, while negative training data are culled in a parallel fashion, though drawn from an alternate dataset. During training, the proposed siamese network conducts binary classification via cross-entropy loss. Subsequently, during the testing phase, we directly extract face verification scores from the network's output layer. Experimental results reveal that the proposed unsupervised system delivers a performance on par with a similar but fully supervised baseline.
Authors: Riccardo Marin, Enric Corona, Gerard Pons-Moll
Aligning a template to 3D human point clouds is a long-standing problem crucial for tasks like animation, reconstruction, and enabling supervised learning pipelines. Recent data-driven methods leverage predicted surface correspondences; however, they are not robust to varied poses or distributions. In contrast, industrial solutions often rely on expensive manual annotations or multi-view capturing systems. Recently, neural fields have shown promising results, but their purely data-driven nature lacks geometric awareness, often resulting in a trivial misalignment of the template registration. In this work, we propose two solutions: LoVD, a novel neural field model that predicts the direction towards the localized SMPL vertices on the target surface; and INT, the first self-supervised task dedicated to neural fields that, at test time, refines the backbone, exploiting the target geometry. We combine them into INLoVD, a robust 3D Human body registration pipeline trained on a large MoCap dataset. INLoVD is efficient (takes less than a minute), solidly achieves the state of the art over public benchmarks, and provides unprecedented generalization on out-of-distribution data. We will release code and checkpoints in \url{url}.
Authors: Rasha Alshawi, Md Tamjidul Hoque, Md Meftahul Ferdaus, Mahdi Abdelguerfi, Kendall Niles, Ken Prathak, Joe Tom, Jordan Klein, Murtada Mousa, Johny Javier Lopez
The proposed architecture, Dual Attentive U-Net with Feature Infusion (DAU-FI Net), addresses challenges in semantic segmentation, particularly on multiclass imbalanced datasets with limited samples. DAU-FI Net integrates multiscale spatial-channel attention mechanisms and feature injection to enhance precision in object localization. The core employs a multiscale depth-separable convolution block, capturing localized patterns across scales. This block is complemented by a spatial-channel squeeze and excitation (scSE) attention unit, modeling inter-dependencies between channels and spatial regions in feature maps. Additionally, additive attention gates refine segmentation by connecting encoder-decoder pathways.
To augment the model, engineered features using Gabor filters for textural analysis, Sobel and Canny filters for edge detection are injected guided by semantic masks to expand the feature space strategically. Comprehensive experiments on a challenging sewer pipe and culvert defect dataset and a benchmark dataset validate DAU-FI Net's capabilities. Ablation studies highlight incremental benefits from attention blocks and feature injection. DAU-FI Net achieves state-of-the-art mean Intersection over Union (IoU) of 95.6% and 98.8% on the defect test set and benchmark respectively, surpassing prior methods by 8.9% and 12.6%, respectively. Ablation studies highlight incremental benefits from attention blocks and feature injection. The proposed architecture provides a robust solution, advancing semantic segmentation for multiclass problems with limited training data. Our sewer-culvert defects dataset, featuring pixel-level annotations, opens avenues for further research in this crucial domain. Overall, this work delivers key innovations in architecture, attention, and feature engineering to elevate semantic segmentation efficacy.
Authors: Zeqian Li, Qirui Chen, Tengda Han, Ya Zhang, Yanfeng Wang, Weidi Xie
In this paper, we consider the problem of temporally aligning the video and texts from instructional videos, specifically, given a long-term video, and associated text sentences, our goal is to determine their corresponding timestamps in the video. To this end, we establish a simple, yet strong model that adopts a Transformer-based architecture with all texts as queries, iteratively attending to the visual features, to infer the optimal timestamp. We conduct thorough experiments to investigate: (i) the effect of upgrading ASR systems to reduce errors from speech recognition, (ii) the effect of various visual-textual backbones, ranging from CLIP to S3D, to the more recent InternVideo, (iii) the effect of transforming noisy ASR transcripts into descriptive steps by prompting a large language model (LLM), to summarize the core activities within the ASR transcript as a new training dataset. As a result, our proposed simple model demonstrates superior performance on both narration alignment and procedural step grounding tasks, surpassing existing state-of-the-art methods by a significant margin on three public benchmarks, namely, 9.3% on HT-Step, 3.4% on HTM-Align and 4.7% on CrossTask. We believe the proposed model and dataset with descriptive steps can be treated as a strong baseline for future research in temporal video-text alignment. All codes, models, and the resulting dataset will be publicly released to the research community.
Authors: Senqiao Yang, Jiaming Liu, Ray Zhang, Mingjie Pan, Zoey Guo, Xiaoqi Li, Zehui Chen, Peng Gao, Yandong Guo, Shanghang Zhang
Recently, Large Language Models (LLMs) and Multimodal Large Language Models (MLLMs) have shown promise in instruction following and 2D image understanding. While these models are powerful, they have not yet been developed to comprehend the more challenging 3D physical scenes, especially when it comes to the sparse outdoor LiDAR data. In this paper, we introduce LiDAR-LLM, which takes raw LiDAR data as input and harnesses the remarkable reasoning capabilities of LLMs to gain a comprehensive understanding of outdoor 3D scenes. The central insight of our LiDAR-LLM is the reformulation of 3D outdoor scene cognition as a language modeling problem, encompassing tasks such as 3D captioning, 3D grounding, 3D question answering, etc. Specifically, due to the scarcity of 3D LiDAR-text pairing data, we introduce a three-stage training strategy and generate relevant datasets, progressively aligning the 3D modality with the language embedding space of LLM. Furthermore, we design a View-Aware Transformer (VAT) to connect the 3D encoder with the LLM, which effectively bridges the modality gap and enhances the LLM's spatial orientation comprehension of visual features. Our experiments show that LiDAR-LLM possesses favorable capabilities to comprehend various instructions regarding 3D scenes and engage in complex spatial reasoning. LiDAR-LLM attains a 40.9 BLEU-1 on the 3D captioning task and achieves a 63.1\% classification accuracy and a 14.3\% BEV mIoU on the 3D grounding task. Web page: https://sites.google.com/view/lidar-llm
Authors: Hayk Manukyan, Andranik Sargsyan, Barsegh Atanyan, Zhangyang Wang, Shant Navasardyan, Humphrey Shi
Recent progress in text-guided image inpainting, based on the unprecedented success of text-to-image diffusion models, has led to exceptionally realistic and visually plausible results. However, there is still significant potential for improvement in current text-to-image inpainting models, particularly in better aligning the inpainted area with user prompts and performing high-resolution inpainting. Therefore, in this paper we introduce HD-Painter, a completely training-free approach that accurately follows to prompts and coherently scales to high-resolution image inpainting. To this end, we design the Prompt-Aware Introverted Attention (PAIntA) layer enhancing self-attention scores by prompt information and resulting in better text alignment generations. To further improve the prompt coherence we introduce the Reweighting Attention Score Guidance (RASG) mechanism seamlessly integrating a post-hoc sampling strategy into general form of DDIM to prevent out-of-distribution latent shifts. Moreover, HD-Painter allows extension to larger scales by introducing a specialized super-resolution technique customized for inpainting, enabling the completion of missing regions in images of up to 2K resolution. Our experiments demonstrate that HD-Painter surpasses existing state-of-the-art approaches qualitatively and quantitatively, achieving an impressive generation accuracy improvement of 61.4% vs 51.9%. We will make the codes publicly available at: https://github.com/Picsart-AI-Research/HD-Painter
Authors: Ana-Maria Marcu, Long Chen, Jan Hünermann, Alice Karnsund, Benoit Hanotte, Prajwal Chidananda, Saurabh Nair, Vijay Badrinarayanan, Alex Kendall, Jamie Shotton, Oleg Sinavski
Autonomous driving has long faced a challenge with public acceptance due to the lack of explainability in the decision-making process. Video question-answering (QA) in natural language provides the opportunity for bridging this gap. Nonetheless, evaluating the performance of Video QA models has proved particularly tough due to the absence of comprehensive benchmarks. To fill this gap, we introduce LingoQA, a benchmark specifically for autonomous driving Video QA. The LingoQA trainable metric demonstrates a 0.95 Spearman correlation coefficient with human evaluations. We introduce a Video QA dataset of central London consisting of 419k samples that we release with the paper. We establish a baseline vision-language model and run extensive ablation studies to understand its performance.
Authors: Philipp Schröppel, Christopher Wewer, Jan Eric Lenssen, Eddy Ilg, Thomas Brox
Controllable generation of 3D assets is important for many practical applications like content creation in movies, games and engineering, as well as in AR/VR. Recently, diffusion models have shown remarkable results in generation quality of 3D objects. However, none of the existing models enable disentangled generation to control the shape and appearance separately. For the first time, we present a suitable representation for 3D diffusion models to enable such disentanglement by introducing a hybrid point cloud and neural radiance field approach. We model a diffusion process over point positions jointly with a high-dimensional feature space for a local density and radiance decoder. While the point positions represent the coarse shape of the object, the point features allow modeling the geometry and appearance details. This disentanglement enables us to sample both independently and therefore to control both separately. Our approach sets a new state of the art in generation compared to previous disentanglement-capable methods by reduced FID scores of 30-90% and is on-par with other non disentanglement-capable state-of-the art methods.
Authors: Dan Kondratyuk, Lijun Yu, Xiuye Gu, José Lezama, Jonathan Huang, Rachel Hornung, Hartwig Adam, Hassan Akbari, Yair Alon, Vighnesh Birodkar, Yong Cheng, Ming-Chang Chiu, Josh Dillon, Irfan Essa, Agrim Gupta, Meera Hahn, Anja Hauth, David Hendon, Alonso Martinez, David Minnen, David Ross, Grant Schindler, Mikhail Sirotenko, Kihyuk Sohn, Krishna Somandepalli, Huisheng Wang, Jimmy Yan, Ming-Hsuan Yang, Xuan Yang, Bryan Seybold, Lu Jiang
We present VideoPoet, a language model capable of synthesizing high-quality video, with matching audio, from a large variety of conditioning signals. VideoPoet employs a decoder-only transformer architecture that processes multimodal inputs -- including images, videos, text, and audio. The training protocol follows that of Large Language Models (LLMs), consisting of two stages: pretraining and task-specific adaptation. During pretraining, VideoPoet incorporates a mixture of multimodal generative objectives within an autoregressive Transformer framework. The pretrained LLM serves as a foundation that can be adapted for a range of video generation tasks. We present empirical results demonstrating the model's state-of-the-art capabilities in zero-shot video generation, specifically highlighting VideoPoet's ability to generate high-fidelity motions. Project page: this http URL
Authors: Bardia Safaei, Vibashan VS, Celso M. de Melo, Vishal M. Patel
Active Learning (AL) aims to enhance the performance of deep models by selecting the most informative samples for annotation from a pool of unlabeled data. Despite impressive performance in closed-set settings, most AL methods fail in real-world scenarios where the unlabeled data contains unknown categories. Recently, a few studies have attempted to tackle the AL problem for the open-set setting. However, these methods focus more on selecting known samples and do not efficiently utilize unknown samples obtained during AL rounds. In this work, we propose an Entropic Open-set AL (EOAL) framework which leverages both known and unknown distributions effectively to select informative samples during AL rounds. Specifically, our approach employs two different entropy scores. One measures the uncertainty of a sample with respect to the known-class distributions. The other measures the uncertainty of the sample with respect to the unknown-class distributions. By utilizing these two entropy scores we effectively separate the known and unknown samples from the unlabeled data resulting in better sampling. Through extensive experiments, we show that the proposed method outperforms existing state-of-the-art methods on CIFAR-10, CIFAR-100, and TinyImageNet datasets. Code is available at \url{https://github.com/bardisafa/EOAL}.
Authors: Shuzhe Wang, Vincent Leroy, Yohann Cabon, Boris Chidlovskii, Jerome Revaud
Multi-view stereo reconstruction (MVS) in the wild requires to first estimate the camera parameters e.g. intrinsic and extrinsic parameters. These are usually tedious and cumbersome to obtain, yet they are mandatory to triangulate corresponding pixels in 3D space, which is the core of all best performing MVS algorithms. In this work, we take an opposite stance and introduce DUSt3R, a radically novel paradigm for Dense and Unconstrained Stereo 3D Reconstruction of arbitrary image collections, i.e. operating without prior information about camera calibration nor viewpoint poses. We cast the pairwise reconstruction problem as a regression of pointmaps, relaxing the hard constraints of usual projective camera models. We show that this formulation smoothly unifies the monocular and binocular reconstruction cases. In the case where more than two images are provided, we further propose a simple yet effective global alignment strategy that expresses all pairwise pointmaps in a common reference frame. We base our network architecture on standard Transformer encoders and decoders, allowing us to leverage powerful pretrained models. Our formulation directly provides a 3D model of the scene as well as depth information, but interestingly, we can seamlessly recover from it, pixel matches, relative and absolute camera. Exhaustive experiments on all these tasks showcase that the proposed DUSt3R can unify various 3D vision tasks and set new SoTAs on monocular/multi-view depth estimation as well as relative pose estimation. In summary, DUSt3R makes many geometric 3D vision tasks easy.
Authors: Tao Huang, Guangqi Jiang, Yanjie Ze, Huazhe Xu
Learning rewards from expert videos offers an affordable and effective solution to specify the intended behaviors for reinforcement learning tasks. In this work, we propose Diffusion Reward, a novel framework that learns rewards from expert videos via conditional video diffusion models for solving complex visual RL problems. Our key insight is that lower generative diversity is observed when conditioned on expert trajectories. Diffusion Reward is accordingly formalized by the negative of conditional entropy that encourages productive exploration of expert-like behaviors. We show the efficacy of our method over 10 robotic manipulation tasks from MetaWorld and Adroit with visual input and sparse reward. Moreover, Diffusion Reward could even solve unseen tasks successfully and effectively, largely surpassing baseline methods. Project page and code: https://diffusion-reward.github.io/.
Authors: Penghao Wu, Saining Xie
When we look around and perform complex tasks, how we see and selectively process what we see is crucial. However, the lack of this visual search mechanism in current multimodal LLMs (MLLMs) hinders their ability to focus on important visual details, especially when handling high-resolution and visually crowded images. To address this, we introduce $\textit{V}^*$, an LLM-guided visual search mechanism that employs the world knowledge in LLMs for efficient visual querying. When combined with an MLLM, this mechanism enhances collaborative reasoning, contextual understanding, and precise targeting of specific visual elements. This integration results in a new MLLM meta-architecture, named $\textbf{S}$how, s$\textbf{EA}$rch, and Tel$\textbf{L}$ (SEAL). We further create $\textit{V}^*$Bench, a benchmark specifically designed to evaluate MLLMs in their ability to process high-resolution images and focus on visual details. Our study highlights the necessity of incorporating visual search capabilities into multimodal systems. The code is available https://github.com/penghao-wu/vstar.
Authors: Qinying Liu, Zilei Wang, Shenghai Rong, Junjie Li, Yixin Zhang
Weakly-supervised temporal action localization aims to localize action instances in videos with only video-level action labels. Existing methods mainly embrace a localization-by-classification pipeline that optimizes the snippet-level prediction with a video classification loss. However, this formulation suffers from the discrepancy between classification and detection, resulting in inaccurate separation of foreground and background (F\&B) snippets. To alleviate this problem, we propose to explore the underlying structure among the snippets by resorting to unsupervised snippet clustering, rather than heavily relying on the video classification loss. Specifically, we propose a novel clustering-based F\&B separation algorithm. It comprises two core components: a snippet clustering component that groups the snippets into multiple latent clusters and a cluster classification component that further classifies the cluster as foreground or background. As there are no ground-truth labels to train these two components, we introduce a unified self-labeling mechanism based on optimal transport to produce high-quality pseudo-labels that match several plausible prior distributions. This ensures that the cluster assignments of the snippets can be accurately associated with their F\&B labels, thereby boosting the F\&B separation. We evaluate our method on three benchmarks: THUMOS14, ActivityNet v1.2 and v1.3. Our method achieves promising performance on all three benchmarks while being significantly more lightweight than previous methods. Code is available at https://github.com/Qinying-Liu/CASE
Authors: Artem Sevastopolsky, Philip-William Grassal, Simon Giebenhain, ShahRukh Athar, Luisa Verdoliva, Matthias Niessner
Current advances in human head modeling allow to generate plausible-looking 3D head models via neural representations. Nevertheless, constructing complete high-fidelity head models with explicitly controlled animation remains an issue. Furthermore, completing the head geometry based on a partial observation, e.g. coming from a depth sensor, while preserving details is often problematic for the existing methods. We introduce a generative model for detailed 3D head meshes on top of an articulated 3DMM which allows explicit animation and high-detail preservation at the same time. Our method is trained in two stages. First, we register a parametric head model with vertex displacements to each mesh of the recently introduced NPHM dataset of accurate 3D head scans. The estimated displacements are baked into a hand-crafted UV layout. Second, we train a StyleGAN model in order to generalize over the UV maps of displacements. The decomposition of the parametric model and high-quality vertex displacements allows us to animate the model and modify it semantically. We demonstrate the results of unconditional generation and fitting to the full or partial observation. The project page is available at https://seva100.github.io/headcraft.
Authors: Qinying Liu, Kecheng Zheng, Wu Wei, Zhan Tong, Yu Liu, Wei Chen, Zilei Wang, Yujun Shen
The crux of learning vision-language models is to extract semantically aligned information from visual and linguistic data. Existing attempts usually face the problem of coarse alignment, \textit{e.g.}, the vision encoder struggles in localizing an attribute-specified object. In this work, we propose an embarrassingly simple approach to better align image and text features with no need of additional data formats other than image-text pairs. Concretely, given an image and its paired text, we manage to parse objects (\textit{e.g.}, cat) and attributes (\textit{e.g.}, black) from the description, which are highly likely to exist in the image. It is noteworthy that the parsing pipeline is fully automatic and thus enjoys good scalability. With these parsed semantics as supervision signals, we can complement the commonly used image-text contrastive loss with the multi-tag classification loss. Extensive experimental results on a broad suite of semantic segmentation datasets substantiate the average 3.65\% improvement of our framework over existing alternatives. Furthermore, the visualization results indicate that attribute supervision makes vision-language models accurately localize attribute-specified objects. Project page can be found at https://qinying-liu.github.io/Tag-Align/
Authors: Chonghao Sima, Katrin Renz, Kashyap Chitta, Li Chen, Hanxue Zhang, Chengen Xie, Ping Luo, Andreas Geiger, Hongyang Li
We study how vision-language models (VLMs) trained on web-scale data can be integrated into end-to-end driving systems to boost generalization and enable interactivity with human users. While recent approaches adapt VLMs to driving via single-round visual question answering (VQA), human drivers reason about decisions in multiple steps. Starting from the localization of key objects, humans estimate object interactions before taking actions. The key insight is that with our proposed task, Graph VQA, where we model graph-structured reasoning through perception, prediction and planning question-answer pairs, we obtain a suitable proxy task to mimic the human reasoning process. We instantiate datasets (DriveLM-Data) built upon nuScenes and CARLA, and propose a VLM-based baseline approach (DriveLM-Agent) for jointly performing Graph VQA and end-to-end driving. The experiments demonstrate that Graph VQA provides a simple, principled framework for reasoning about a driving scene, and DriveLM-Data provides a challenging benchmark for this task. Our DriveLM-Agent baseline performs end-to-end autonomous driving competitively in comparison to state-of-the-art driving-specific architectures. Notably, its benefits are pronounced when it is evaluated zero-shot on unseen objects or sensor configurations. We hope this work can be the starting point to shed new light on how to apply VLMs for autonomous driving. To facilitate future research, all code, data, and models are available to the public.
Authors: Yen-Chi Cheng, Chieh Hubert Lin, Chaoyang Wang, Yash Kant, Sergey Tulyakov, Alexander Schwing, Liangyan Gui, Hsin-Ying Lee
Toward unlocking the potential of generative models in immersive 4D experiences, we introduce Virtual Pet, a novel pipeline to model realistic and diverse motions for target animal species within a 3D environment. To circumvent the limited availability of 3D motion data aligned with environmental geometry, we leverage monocular internet videos and extract deformable NeRF representations for the foreground and static NeRF representations for the background. For this, we develop a reconstruction strategy, encompassing species-level shared template learning and per-video fine-tuning. Utilizing the reconstructed data, we then train a conditional 3D motion model to learn the trajectory and articulation of foreground animals in the context of 3D backgrounds. We showcase the efficacy of our pipeline with comprehensive qualitative and quantitative evaluations using cat videos. We also demonstrate versatility across unseen cats and indoor environments, producing temporally coherent 4D outputs for enriched virtual experiences.
Authors: Christen Millerdurai, Diogo Luvizon, Viktor Rudnev, André Jonas, Jiayi Wang, Christian Theobalt, Vladislav Golyanik
3D hand tracking from a monocular video is a very challenging problem due to hand interactions, occlusions, left-right hand ambiguity, and fast motion. Most existing methods rely on RGB inputs, which have severe limitations under low-light conditions and suffer from motion blur. In contrast, event cameras capture local brightness changes instead of full image frames and do not suffer from the described effects. Unfortunately, existing image-based techniques cannot be directly applied to events due to significant differences in the data modalities. In response to these challenges, this paper introduces the first framework for 3D tracking of two fast-moving and interacting hands from a single monocular event camera. Our approach tackles the left-right hand ambiguity with a novel semi-supervised feature-wise attention mechanism and integrates an intersection loss to fix hand collisions. To facilitate advances in this research domain, we release a new synthetic large-scale dataset of two interacting hands, Ev2Hands-S, and a new real benchmark with real event streams and ground-truth 3D annotations, Ev2Hands-R. Our approach outperforms existing methods in terms of the 3D reconstruction accuracy and generalises to real data under severe light conditions.
Authors: Ken C. L. Wong, Satyananda Kashyap, Mehdi Moradi
Network-based transfer learning allows the reuse of deep learning features with limited data, but the resulting models can be unnecessarily large. Although network pruning can improve inference efficiency, existing algorithms usually require fine-tuning that may not be suitable for small datasets. In this paper, using the singular value decomposition, we decompose a convolutional layer into two layers: a convolutional layer with the orthonormal basis vectors as the filters, and a "BasisScalingConv" layer which is responsible for rescaling the features and transforming them back to the original space. As the filters in each decomposed layer are linearly independent, when using the proposed basis scaling factors with the Taylor approximation of importance, pruning can be more effective and fine-tuning individual weights is unnecessary. Furthermore, as the numbers of input and output channels of the original convolutional layer remain unchanged after basis pruning, it is applicable to virtually all architectures and can be combined with existing pruning algorithms for double pruning to further increase the pruning capability. When transferring knowledge from ImageNet pre-trained models to different target domains, with less than 1% reduction in classification accuracies, we can achieve pruning ratios up to 74.6% for CIFAR-10 and 98.9% for MNIST in model parameters.
Authors: Xinzhu Ma, Wanli Ouyang, Andrea Simonelli, Elisa Ricci
3D object detection from images, one of the fundamental and challenging problems in autonomous driving, has received increasing attention from both industry and academia in recent years. Benefiting from the rapid development of deep learning technologies, image-based 3D detection has achieved remarkable progress. Particularly, more than 200 works have studied this problem from 2015 to 2021, encompassing a broad spectrum of theories, algorithms, and applications. However, to date no recent survey exists to collect and organize this knowledge. In this paper, we fill this gap in the literature and provide the first comprehensive survey of this novel and continuously growing research field, summarizing the most commonly used pipelines for image-based 3D detection and deeply analyzing each of their components. Additionally, we also propose two new taxonomies to organize the state-of-the-art methods into different categories, with the intent of providing a more systematic review of existing methods and facilitating fair comparisons with future works. In retrospect of what has been achieved so far, we also analyze the current challenges in the field and discuss future directions for image-based 3D detection research.
Authors: Qinying Liu, Zilei Wang, Ruoxi Chen, Zhilin Li
Weakly-supervised temporal action localization (WTAL) intends to detect action instances with only weak supervision, \eg, video-level labels. The current~\textit{de facto} pipeline locates action instances by thresholding and grouping continuous high-score regions on temporal class activation sequences. In this route, the capacity of the model to recognize the relationships between adjacent snippets is of vital importance which determines the quality of the action boundaries. However, it is error-prone since the variations between adjacent snippets are typically subtle, and unfortunately this is overlooked in the literature. To tackle the issue, we propose a novel WTAL approach named Convex Combination Consistency between Neighbors (C$^3$BN). C$^3$BN consists of two key ingredients: a micro data augmentation strategy that increases the diversity in-between adjacent snippets by convex combination of adjacent snippets, and a macro-micro consistency regularization that enforces the model to be invariant to the transformations~\textit{w.r.t.} video semantics, snippet predictions, and snippet representations. Consequently, fine-grained patterns in-between adjacent snippets are enforced to be explored, thereby resulting in a more robust action boundary localization. Experimental results demonstrate the effectiveness of C$^3$BN on top of various baselines for WTAL with video-level and point-level supervisions. Code is at https://github.com/Qinying-Liu/C3BN.
Authors: Víctor Corsino, José Manuel Gilpérez, Luis Herrera
The resolution of intelligence tests, in particular numerical sequences, has been of great interest in the evaluation of AI systems. We present a new computational model called KitBit that uses a reduced set of algorithms and their combinations to build a predictive model that finds the underlying pattern in numerical sequences, such as those included in IQ tests and others of much greater complexity. We present the fundamentals of the model and its application in different cases. First, the system is tested on a set of number series used in IQ tests collected from various sources. Next, our model is successfully applied on the sequences used to evaluate the models reported in the literature. In both cases, the system is capable of solving these types of problems in less than a second using standard computing power. Finally, KitBit's algorithms have been applied for the first time to the complete set of entire sequences of the well-known OEIS database. We find a pattern in the form of a list of algorithms and predict the following terms in the largest number of series to date. These results demonstrate the potential of KitBit to solve complex problems that could be represented numerically.
Authors: Md. Iqbal Hossain, Mohammad Zunaed, Md. Kawsar Ahmed, S. M. Jawwad Hossain, Anwarul Hasan, Taufiq Hasan
Objective: Computer-aided disease diagnosis and prognosis based on medical images is a rapidly emerging field. Many Convolutional Neural Network (CNN) architectures have been developed by researchers for disease classification and localization from chest X-ray images. It is known that different thoracic disease lesions are more likely to occur in specific anatomical regions compared to others. This article aims to incorporate this disease and region-dependent prior probability distribution within a deep learning framework. Methods: We present the ThoraX-PriorNet, a novel attention-based CNN model for thoracic disease classification. We first estimate a disease-dependent spatial probability, i.e., an anatomical prior, that indicates the probability of occurrence of a disease in a specific region in a chest X-ray image. Next, we develop a novel attention-based classification model that combines information from the estimated anatomical prior and automatically extracted chest region of interest (ROI) masks to provide attention to the feature maps generated from a deep convolution network. Unlike previous works that utilize various self-attention mechanisms, the proposed method leverages the extracted chest ROI masks along with the probabilistic anatomical prior information, which selects the region of interest for different diseases to provide attention. Results: The proposed method shows superior performance in disease classification on the NIH ChestX-ray14 dataset compared to existing state-of-the-art methods while reaching an area under the ROC curve (%AUC) of 84.67. Regarding disease localization, the anatomy prior attention method shows competitive performance compared to state-of-the-art methods, achieving an accuracy of 0.80, 0.63, 0.49, 0.33, 0.28, 0.21, and 0.04 with an Intersection over Union (IoU) threshold of 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, and 0.7, respectively.
Authors: Zhaoyuan Yang, Zhiwei Xu, Jing Zhang, Richard Hartley, Peter Tu
In this work, we formulate a novel framework for adversarial robustness using the manifold hypothesis. This framework provides sufficient conditions for defending against adversarial examples. We develop an adversarial purification method with this framework. Our method combines manifold learning with variational inference to provide adversarial robustness without the need for expensive adversarial training. Experimentally, our approach can provide adversarial robustness even if attackers are aware of the existence of the defense. In addition, our method can also serve as a test-time defense mechanism for variational autoencoders.
Authors: Rihao Chang, Yongtao Ma, Tong Hao, Weizhi Nie
The surge in 3D modeling has led to a pronounced research emphasis on the field of 3D shape retrieval. Numerous contemporary approaches have been put forth to tackle this intricate challenge. Nevertheless, effectively addressing the intricacies of cross-modal 3D shape retrieval remains a formidable undertaking, owing to inherent modality-based disparities. This study presents an innovative notion, termed "geometric words", which functions as elemental constituents for representing entities through combinations. To establish the knowledge graph, we employ geometric words as nodes, connecting them via shape categories and geometry attributes. Subsequently, we devise a unique graph embedding method for knowledge acquisition. Finally, an effective similarity measure is introduced for retrieval purposes. Importantly, each 3D or 2D entity can anchor its geometric terms within the knowledge graph, thereby serving as a link between cross-domain data. As a result, our approach facilitates multiple cross-domain 3D shape retrieval tasks. We evaluate the proposed method's performance on the ModelNet40 and ShapeNetCore55 datasets, encompassing scenarios related to 3D shape retrieval and cross-domain retrieval. Furthermore, we employ the established cross-modal dataset (MI3DOR) to assess cross-modal 3D shape retrieval. The resulting experimental outcomes, in conjunction with comparisons against state-of-the-art techniques, clearly highlight the superiority of our approach.
Authors: Shangchao Su, Mingzhao Yang, Bin Li, Xiangyang Xue
Federated learning (FL) enables multiple clients to collaboratively train a global model without disclosing their data. Previous researches often require training the complete model parameters. However, the emergence of powerful pre-trained models makes it possible to achieve higher performance with fewer learnable parameters in FL. In this paper, we propose a federated adaptive prompt tuning algorithm, FedAPT, for multi-domain collaborative image classification with powerful foundation models, like CLIP. Compared with direct federated prompt tuning, our core idea is to adaptively unlock specific domain knowledge for each test sample in order to provide them with personalized prompts. To implement this idea, we design an adaptive prompt tuning module, which consists of a meta prompt, an adaptive network, and some keys. The server randomly generates a set of keys and assigns a unique key to each client. Then all clients cooperatively train the global adaptive network and meta prompt with the local datasets and the frozen keys. Ultimately, the global aggregation model can assign a personalized prompt to CLIP based on the domain features of each test sample. We perform extensive experiments on two multi-domain image classification datasets across two different settings -- supervised and unsupervised. The results show that FedAPT can achieve better performance with less than 10\% of the number of parameters of the fully trained model, and the global model can perform well in diverse client domains simultaneously.
Authors: Zeyu Shangguan, Lian Huai, Tong Liu, Xingqun Jiang
Due to the scarcity of sampling data in reality, few-shot object detection (FSOD) has drawn more and more attention because of its ability to quickly train new detection concepts with less data. However, there are still failure identifications due to the difficulty in distinguishing confusable classes. We also notice that the high standard deviation of average precision reveals the inconsistent detection performance. To this end, we propose a novel FSOD method with Refined Contrastive Learning (FSRC). A pre-determination component is introduced to find out the Resemblance Group from novel classes which contains confusable classes. Afterwards, Refined Contrastive Learning (RCL) is pointedly performed on this group of classes in order to increase the inter-class distances among them. In the meantime, the detection results distribute more uniformly which further improve the performance. Experimental results based on PASCAL VOC and COCO datasets demonstrate our proposed method outperforms the current state-of-the-art research.
Authors: YuTeng Ye, Hang Zhou, Jiale Cai, Chenxing Gao, Youjia Zhang, Junle Wang, Qiang Hu, Junqing Yu, Wei Yang
Occluded person re-identification (ReID) is a challenging problem due to contamination from occluders. Existing approaches address the issue with prior knowledge cues, such as human body key points and semantic segmentations, which easily fail in the presence of heavy occlusion and other humans as occluders. In this paper, we propose a feature pruning and consolidation (FPC) framework to circumvent explicit human structure parsing. The framework mainly consists of a sparse encoder, a multi-view feature mathcing module, and a feature consolidation decoder. Specifically, the sparse encoder drops less important image tokens, mostly related to background noise and occluders, solely based on correlation within the class token attention. Subsequently, the matching stage relies on the preserved tokens produced by the sparse encoder to identify k-nearest neighbors in the gallery by measuring the image and patch-level combined similarity. Finally, we use the feature consolidation module to compensate pruned features using identified neighbors for recovering essential information while disregarding disturbance from noise and occlusion. Experimental results demonstrate the effectiveness of our proposed framework on occluded, partial, and holistic Re-ID datasets. In particular, our method outperforms state-of-the-art results by at least 8.6\% mAP and 6.0\% Rank-1 accuracy on the challenging Occluded-Duke dataset.
Authors: Pratik K. Mishra, Alex Mihailidis, Shehroz S. Khan
The existing methods for video anomaly detection mostly utilize videos containing identifiable facial and appearance-based features. The use of videos with identifiable faces raises privacy concerns, especially when used in a hospital or community-based setting. Appearance-based features can also be sensitive to pixel-based noise, straining the anomaly detection methods to model the changes in the background and making it difficult to focus on the actions of humans in the foreground. Structural information in the form of skeletons describing the human motion in the videos is privacy-protecting and can overcome some of the problems posed by appearance-based features. In this paper, we present a survey of privacy-protecting deep learning anomaly detection methods using skeletons extracted from videos. We present a novel taxonomy of algorithms based on the various learning approaches. We conclude that skeleton-based approaches for anomaly detection can be a plausible privacy-protecting alternative for video anomaly detection. Lastly, we identify major open research questions and provide guidelines to address them.
Authors: Tsz Chung Wong, Abubakar Sani-Mohammed, Jinhong Wang, Puzuo Wang, Wei Yao, Marco Heurich
Understanding forest health is of great importance for the conservation of the integrity of forest ecosystems. In this regard, evaluating the amount and quality of dead wood is of utmost interest as they are favorable indicators of biodiversity. Apparently, remote sensing-based machine learning techniques have proven to be more efficient and sustainable with unprecedented accuracy in forest inventory. This study, for the first time, automatically categorizing individual coniferous trees (Norway spruce) into five decay stages (live, declining, dead, loose bark, and clean) from combined airborne laser scanning (ALS) point clouds and color infrared (CIR) images using three different Machine Learning methods - 3D point cloud-based deep learning (KPConv), Convolutional Neural Network (CNN), and Random Forest (RF). First, CIR colorized point clouds are created by fusing the ALS point clouds and color infrared images. Then, individual tree segmentation is conducted, after which the results are further projected onto four orthogonal planes. Finally, the classification is conducted on the two datasets (3D multispectral point clouds and 2D projected images) based on the three Machine Learning algorithms. All models achieved promising results, reaching overall accuracy (OA) of up to 88.8%, 88.4% and 85.9% for KPConv, CNN and RF, respectively. The experimental results reveal that color information, 3D coordinates, and intensity of point clouds have significant impact on the promising classification performance. The performance of our models, therefore, shows the significance of machine/deep learning for individual tree decay stages classification and landscape-wide assessment of the dead wood amount and quality by using modern airborne remote sensing techniques. The proposed method can contribute as an important and reliable tool for monitoring biodiversity in forest ecosystems.
Authors: Wei-Yin Ko, Daniel D'souza, Karina Nguyen, Randall Balestriero, Sara Hooker
Ensembling multiple Deep Neural Networks (DNNs) is a simple and effective way to improve top-line metrics and to outperform a larger single model. In this work, we go beyond top-line metrics and instead explore the impact of ensembling on subgroup performances. Surprisingly, we observe that even with a simple homogeneous ensemble -- all the individual DNNs share the same training set, architecture, and design choices -- the minority group performance disproportionately improves with the number of models compared to the majority group, i.e. fairness naturally emerges from ensembling. Even more surprising, we find that this gain keeps occurring even when a large number of models is considered, e.g. $20$, despite the fact that the average performance of the ensemble plateaus with fewer models. Our work establishes that simple DNN ensembles can be a powerful tool for alleviating disparate impact from DNN classifiers, thus curbing algorithmic harm. We also explore why this is the case. We find that even in homogeneous ensembles, varying the sources of stochasticity through parameter initialization, mini-batch sampling, and data-augmentation realizations, results in different fairness outcomes.
Authors: Marin Scalbert, Maria Vakalopoulou, Florent Couzinié-Devy
In Self-Supervised Learning (SSL), models are typically pretrained, fine-tuned, and evaluated on the same domains. However, they tend to perform poorly when evaluated on unseen domains, a challenge that Unsupervised Domain Generalization (UDG) seeks to address. Current UDG methods rely on domain labels, which are often challenging to collect, and domain-specific architectures that lack scalability when confronted with numerous domains, making the current methodology impractical and rigid. Inspired by contrastive-based UDG methods that mitigate spurious correlations by restricting comparisons to examples from the same domain, we hypothesize that eliminating style variability within a batch could provide a more convenient and flexible way to reduce spurious correlations without requiring domain labels. To verify this hypothesis, we introduce Batch Styles Standardization (BSS), a relatively simple yet powerful Fourier-based method to standardize the style of images in a batch specifically designed for integration with SSL methods to tackle UDG. Combining BSS with existing SSL methods offers serious advantages over prior UDG methods: (1) It eliminates the need for domain labels or domain-specific network components to enhance domain-invariance in SSL representations, and (2) offers flexibility as BSS can be seamlessly integrated with diverse contrastive-based but also non-contrastive-based SSL methods. Experiments on several UDG datasets demonstrate that it significantly improves downstream task performances on unseen domains, often outperforming or rivaling with UDG methods. Finally, this work clarifies the underlying mechanisms contributing to BSS's effectiveness in improving domain-invariance in SSL representations and performances on unseen domain.
Authors: Yusuke Hirota, Yuta Nakashima, Noa Garcia
Image captioning models are known to perpetuate and amplify harmful societal bias in the training set. In this work, we aim to mitigate such gender bias in image captioning models. While prior work has addressed this problem by forcing models to focus on people to reduce gender misclassification, it conversely generates gender-stereotypical words at the expense of predicting the correct gender. From this observation, we hypothesize that there are two types of gender bias affecting image captioning models: 1) bias that exploits context to predict gender, and 2) bias in the probability of generating certain (often stereotypical) words because of gender. To mitigate both types of gender biases, we propose a framework, called LIBRA, that learns from synthetically biased samples to decrease both types of biases, correcting gender misclassification and changing gender-stereotypical words to more neutral ones. Code is available at https://github.com/rebnej/LIBRA.
Authors: Chuanfei Hu, Tianyi Xia, Shenghong Ju, Xinde Li
Learning to segmentation without large-scale samples is an inherent capability of human. Recently, Segment Anything Model (SAM) performs the significant zero-shot image segmentation, attracting considerable attention from the computer vision community. Here, we investigate the capability of SAM for medical image analysis, especially for multi-phase liver tumor segmentation (MPLiTS), in terms of prompts, data resolution, phases. Experimental results demonstrate that there might be a large gap between SAM and expected performance. Fortunately, the qualitative results show that SAM is a powerful annotation tool for the community of interactive medical image segmentation.
Authors: Xiangtai Li, Henghui Ding, Haobo Yuan, Wenwei Zhang, Jiangmiao Pang, Guangliang Cheng, Kai Chen, Ziwei Liu, Chen Change Loy
Visual segmentation seeks to partition images, video frames, or point clouds into multiple segments or groups. This technique has numerous real-world applications, such as autonomous driving, image editing, robot sensing, and medical analysis. Over the past decade, deep learning-based methods have made remarkable strides in this area. Recently, transformers, a type of neural network based on self-attention originally designed for natural language processing, have considerably surpassed previous convolutional or recurrent approaches in various vision processing tasks. Specifically, vision transformers offer robust, unified, and even simpler solutions for various segmentation tasks. This survey provides a thorough overview of transformer-based visual segmentation, summarizing recent advancements. We first review the background, encompassing problem definitions, datasets, and prior convolutional methods. Next, we summarize a meta-architecture that unifies all recent transformer-based approaches. Based on this meta-architecture, we examine various method designs, including modifications to the meta-architecture and associated applications. We also present several closely related settings, including 3D point cloud segmentation, foundation model tuning, domain-aware segmentation, efficient segmentation, and medical segmentation. Additionally, we compile and re-evaluate the reviewed methods on several well-established datasets. Finally, we identify open challenges in this field and propose directions for future research. The project page can be found at https://github.com/lxtGH/Awesome-Segmentation-With-Transformer. We will also continually monitor developments in this rapidly evolving field.
Authors: Ekta U. Samani, Ashis G. Banerjee
Recognition of occluded objects in unseen and unstructured indoor environments is a challenging problem for mobile robots. To address this challenge, we propose a new descriptor, TOPS, for point clouds generated from depth images and an accompanying recognition framework, THOR, inspired by human reasoning. The descriptor employs a novel slicing-based approach to compute topological features from filtrations of simplicial complexes using persistent homology, and facilitates reasoning-based recognition using object unity. Apart from a benchmark dataset, we report performance on a new dataset, the UW Indoor Scenes (UW-IS) Occluded dataset, curated using commodity hardware to reflect real-world scenarios with different environmental conditions and degrees of object occlusion. THOR outperforms state-of-the-art methods on both the datasets and achieves substantially higher recognition accuracy for all the scenarios of the UW-IS Occluded dataset. Therefore, THOR, is a promising step toward robust recognition in low-cost robots, meant for everyday use in indoor settings.
Authors: Teerapong Panboonyuen, Naphat Nithisopa, Panin Pienroj, Laphonchai Jirachuphun, Chaiwasut Watthanasirikrit, Naruepon Pornwiriyakul
Evaluating car damages from misfortune is critical to the car insurance industry. However, the accuracy is still insufficient for real-world applications since the deep learning network is not designed for car damage images as inputs, and its segmented masks are still very coarse. This paper presents MARS (Mask Attention Refinement with Sequential quadtree nodes) for car damage instance segmentation. Our MARS represents self-attention mechanisms to draw global dependencies between the sequential quadtree nodes layer and quadtree transformer to recalibrate channel weights and predict highly accurate instance masks. Our extensive experiments demonstrate that MARS outperforms state-of-the-art (SOTA) instance segmentation methods on three popular benchmarks such as Mask R-CNN [9], PointRend [13], and Mask Transfiner [12], by a large margin of +1.3 maskAP-based R50-FPN backbone and +2.3 maskAP-based R101-FPN backbone on Thai car-damage dataset. Our demos are available at https://github.com/kaopanboonyuen/MARS.
Authors: Alceu Bissoto, Catarina Barata, Eduardo Valle, Sandra Avila
Distribution shifts are common in real-world datasets and can affect the performance and reliability of deep learning models. In this paper, we study two types of distribution shifts: diversity shifts, which occur when test samples exhibit patterns unseen during training, and correlation shifts, which occur when test data present a different correlation between seen invariant and spurious features. We propose an integrated protocol to analyze both types of shifts using datasets where they co-exist in a controllable manner. Finally, we apply our approach to a real-world classification problem of skin cancer analysis, using out-of-distribution datasets and specialized bias annotations. Our protocol reveals three findings: 1) Models learn and propagate correlation shifts even with low-bias training; this poses a risk of accumulating and combining unaccountable weak biases; 2) Models learn robust features in high- and low-bias scenarios but use spurious ones if test samples have them; this suggests that spurious correlations do not impair the learning of robust features; 3) Diversity shift can reduce the reliance on spurious correlations; this is counter intuitive since we expect biased models to depend more on biases when invariant features are missing. Our work has implications for distribution shift research and practice, providing new insights into how models learn and rely on spurious correlations under different types of shifts.
Authors: Junde Wu, Jiayuan Zhu, Yuanpei Liu, Yueming Jin, Min Xu
Large foundation models, known for their strong zero-shot generalization, have excelled in visual and language applications. However, applying them to medical image segmentation, a domain with diverse imaging types and target labels, remains an open challenge. Current approaches, such as adapting interactive segmentation models like Segment Anything Model (SAM), require user prompts for each sample during inference. Alternatively, transfer learning methods like few/one-shot models demand labeled samples, leading to high costs. This paper introduces a new paradigm toward the universal medical image segmentation, termed 'One-Prompt Segmentation.' One-Prompt Segmentation combines the strengths of one-shot and interactive methods. In the inference stage, with just \textbf{one prompted sample}, it can adeptly handle the unseen task in a single forward pass. We train One-Prompt Model on 64 open-source medical datasets, accompanied by the collection of over 3,000 clinician-labeled prompts. Tested on 14 previously unseen tasks, the One-Prompt Model showcases superior zero-shot segmentation capabilities, outperforming a wide range of related methods. The code and annotated data will be publicly released.
Authors: Pengxin Zeng, Mouxing Yang, Yiding Lu, Changqing Zhang, Peng Hu, Xi Peng
Robust multi-view learning with incomplete information has received significant attention due to issues such as incomplete correspondences and incomplete instances that commonly affect real-world multi-view applications. Existing approaches heavily rely on paired samples to realign or impute defective ones, but such preconditions cannot always be satisfied in practice due to the complexity of data collection and transmission. To address this problem, we present a novel framework called SeMantic Invariance LEarning (SMILE) for multi-view clustering with incomplete information that does not require any paired samples. To be specific, we discover the existence of invariant semantic distribution across different views, which enables SMILE to alleviate the cross-view discrepancy to learn consensus semantics without requiring any paired samples. The resulting consensus semantics remain unaffected by cross-view distribution shifts, making them useful for realigning/imputing defective instances and forming clusters. We demonstrate the effectiveness of SMILE through extensive comparison experiments with 13 state-of-the-art baselines on five benchmarks. Our approach improves the clustering accuracy of NoisyMNIST from 19.3\%/23.2\% to 82.7\%/69.0\% when the correspondences/instances are fully incomplete. The code could be accessed from https://pengxi.me.
Authors: Sungnyun Kim, Junsoo Lee, Kibeom Hong, Daesik Kim, Namhyuk Ahn
In this study, we aim to extend the capabilities of diffusion-based text-to-image (T2I) generation models by incorporating diverse modalities beyond textual description, such as sketch, box, color palette, and style embedding, within a single model. We thus design a multimodal T2I diffusion model, coined as DiffBlender, by separating the channels of conditions into three types, i.e., image forms, spatial tokens, and non-spatial tokens. The unique architecture of DiffBlender facilitates adding new input modalities, pioneering a scalable framework for conditional image generation. Notably, we achieve this without altering the parameters of the existing generative model, Stable Diffusion, only with updating partial components. Our study establishes new benchmarks in multimodal generation through quantitative and qualitative comparisons with existing conditional generation methods. We demonstrate that DiffBlender faithfully blends all the provided information and showcase its various applications in the detailed image synthesis.
Authors: Jean-Yves Franceschi, Mike Gartrell, Ludovic Dos Santos, Thibaut Issenhuth, Emmanuel de Bézenac, Mickaël Chen, Alain Rakotomamonjy
Particle-based deep generative models, such as gradient flows and score-based diffusion models, have recently gained traction thanks to their striking performance. Their principle of displacing particle distributions using differential equations is conventionally seen as opposed to the previously widespread generative adversarial networks (GANs), which involve training a pushforward generator network. In this paper we challenge this interpretation, and propose a novel framework that unifies particle and adversarial generative models by framing generator training as a generalization of particle models. This suggests that a generator is an optional addition to any such generative model. Consequently, integrating a generator into a score-based diffusion model and training a GAN without a generator naturally emerge from our framework. We empirically test the viability of these original models as proofs of concepts of potential applications of our framework.
Authors: Xinchao Song, Nikolas Lamb, Sean Banerjee, Natasha Kholgade Banerjee
We present Pix2Repair, an automated shape repair approach that generates restoration shapes from images to repair fractured objects. Prior repair approaches require a high-resolution watertight 3D mesh of the fractured object as input. Input 3D meshes must be obtained using expensive 3D scanners, and scanned meshes require manual cleanup, limiting accessibility and scalability. Pix2Repair takes an image of the fractured object as input and automatically generates a 3D printable restoration shape. We contribute a novel shape function that deconstructs a latent code representing the fractured object into a complete shape and a break surface. We also introduce Fantastic Breaks Imaged, the first large-scale dataset of 11,653 real-world images of fractured objects for training and evaluating image-based shape repair approaches. Our dataset contains images of objects from Fantastic Breaks, complete with rich annotations. We show restorations for real fractures from our dataset, and for synthetic fractures from the Geometric Breaks and Breaking Bad datasets. Our approach outperforms shape completion approaches adapted for shape repair in terms of chamfer distance, normal consistency, and percent restorations generated.
Authors: Zeyue Xue, Guanglu Song, Qiushan Guo, Boxiao Liu, Zhuofan Zong, Yu Liu, Ping Luo
Text-to-image generation has recently witnessed remarkable achievements. We introduce a text-conditional image diffusion model, termed RAPHAEL, to generate highly artistic images, which accurately portray the text prompts, encompassing multiple nouns, adjectives, and verbs. This is achieved by stacking tens of mixture-of-experts (MoEs) layers, i.e., space-MoE and time-MoE layers, enabling billions of diffusion paths (routes) from the network input to the output. Each path intuitively functions as a "painter" for depicting a particular textual concept onto a specified image region at a diffusion timestep. Comprehensive experiments reveal that RAPHAEL outperforms recent cutting-edge models, such as Stable Diffusion, ERNIE-ViLG 2.0, DeepFloyd, and DALL-E 2, in terms of both image quality and aesthetic appeal. Firstly, RAPHAEL exhibits superior performance in switching images across diverse styles, such as Japanese comics, realism, cyberpunk, and ink illustration. Secondly, a single model with three billion parameters, trained on 1,000 A100 GPUs for two months, achieves a state-of-the-art zero-shot FID score of 6.61 on the COCO dataset. Furthermore, RAPHAEL significantly surpasses its counterparts in human evaluation on the ViLG-300 benchmark. We believe that RAPHAEL holds the potential to propel the frontiers of image generation research in both academia and industry, paving the way for future breakthroughs in this rapidly evolving field. More details can be found on a webpage: https://raphael-painter.github.io/.
Authors: Roi Peleg, Teddy Lazebnik, Assaf Hoogi
The performance improvement of deep networks significantly depends on their optimizers. With existing optimizers, precise and efficient recognition of the gradients trend remains a challenge. Existing optimizers predominantly adopt techniques based on the first-order exponential moving average (EMA), which results in noticeable delays that impede the real-time tracking of gradients trend and consequently yield sub-optimal performance. To overcome this limitation, we introduce a novel optimizer called fast-adaptive moment estimation (FAME). Inspired by the triple exponential moving average (TEMA) used in the financial domain, FAME leverages the potency of higher-order TEMA to improve the precision of identifying gradient trends. TEMA plays a central role in the learning process as it actively influences optimization dynamics; this role differs from its conventional passive role as a technical indicator in financial contexts. Because of the introduction of TEMA into the optimization process, FAME can identify gradient trends with higher accuracy and fewer lag issues, thereby offering smoother and more consistent responses to gradient fluctuations compared to conventional first-order EMA. To study the effectiveness of our novel FAME optimizer, we conducted comprehensive experiments encompassing six diverse computer-vision benchmarks and tasks, spanning detection, classification, and semantic comprehension. We integrated FAME into 15 learning architectures and compared its performance with those of six popular optimizers. Results clearly showed that FAME is more robust and accurate and provides superior performance stability by minimizing noise (i.e., trend fluctuations). Notably, FAME achieves higher accuracy levels in remarkably fewer training epochs than its counterparts, clearly indicating its significance for optimizing deep networks in computer-vision tasks.
Authors: Michael Tschannen, Manoj Kumar, Andreas Steiner, Xiaohua Zhai, Neil Houlsby, Lucas Beyer
Contrastive pretraining on image-text pairs from the web is one of the most popular large-scale pretraining strategies for vision backbones, especially in the context of large multimodal models. At the same time, image captioning on this type of data is commonly considered an inferior pretraining strategy. In this paper, we perform a fair comparison of these two pretraining strategies, carefully matching training data, compute, and model capacity. Using a standard encoder-decoder transformer, we find that captioning alone is surprisingly effective: on classification tasks, captioning produces vision encoders competitive with contrastively pretrained encoders, while surpassing them on vision & language tasks. We further analyze the effect of the model architecture and scale, as well as the pretraining data on the representation quality, and find that captioning exhibits the same or better scaling behavior along these axes. Overall our results show that plain image captioning is a more powerful pretraining strategy than was previously believed.
Authors: Denys Rozumnyi, Stefan Popov, Kevis-Kokitsi Maninis, Matthias Nießner, Vittorio Ferrari
Indoor rooms are among the most common use cases in 3D scene understanding. Current state-of-the-art methods for this task are driven by large annotated datasets. Room layouts are especially important, consisting of structural elements in 3D, such as wall, floor, and ceiling. However, they are difficult to annotate, especially on pure RGB video. We propose a novel method to produce generic 3D room layouts just from 2D segmentation masks, which are easy to annotate for humans. Based on these 2D annotations, we automatically reconstruct 3D plane equations for the structural elements and their spatial extent in the scene, and connect adjacent elements at the appropriate contact edges. We annotate and publicly release 2246 3D room layouts on the RealEstate10k dataset, containing YouTube videos. We demonstrate the high quality of these 3D layouts annotations with extensive experiments.
Authors: Xudong Wang, Shufan Li, Konstantinos Kallidromitis, Yusuke Kato, Kazuki Kozuka, Trevor Darrell
Open-vocabulary image segmentation aims to partition an image into semantic regions according to arbitrary text descriptions. However, complex visual scenes can be naturally decomposed into simpler parts and abstracted at multiple levels of granularity, introducing inherent segmentation ambiguity. Unlike existing methods that typically sidestep this ambiguity and treat it as an external factor, our approach actively incorporates a hierarchical representation encompassing different semantic-levels into the learning process. We propose a decoupled text-image fusion mechanism and representation learning modules for both "things" and "stuff". Additionally, we systematically examine the differences that exist in the textual and visual features between these types of categories. Our resulting model, named HIPIE, tackles HIerarchical, oPen-vocabulary, and unIvErsal segmentation tasks within a unified framework. Benchmarked on over 40 datasets, e.g., ADE20K, COCO, Pascal-VOC Part, RefCOCO/RefCOCOg, ODinW and SeginW, HIPIE achieves the state-of-the-art results at various levels of image comprehension, including semantic-level (e.g., semantic segmentation), instance-level (e.g., panoptic/referring segmentation and object detection), as well as part-level (e.g., part/subpart segmentation) tasks. Our code is released at https://github.com/berkeley-hipie/HIPIE.
Authors: Wenhao Tang, Sheng Huang, Xiaoxian Zhang, Fengtao Zhou, Yi Zhang, Bo Liu
The whole slide image (WSI) classification is often formulated as a multiple instance learning (MIL) problem. Since the positive tissue is only a small fraction of the gigapixel WSI, existing MIL methods intuitively focus on identifying salient instances via attention mechanisms. However, this leads to a bias towards easy-to-classify instances while neglecting hard-to-classify instances. Some literature has revealed that hard examples are beneficial for modeling a discriminative boundary accurately. By applying such an idea at the instance level, we elaborate a novel MIL framework with masked hard instance mining (MHIM-MIL), which uses a Siamese structure (Teacher-Student) with a consistency constraint to explore the potential hard instances. With several instance masking strategies based on attention scores, MHIM-MIL employs a momentum teacher to implicitly mine hard instances for training the student model, which can be any attention-based MIL model. This counter-intuitive strategy essentially enables the student to learn a better discriminating boundary. Moreover, the student is used to update the teacher with an exponential moving average (EMA), which in turn identifies new hard instances for subsequent training iterations and stabilizes the optimization. Experimental results on the CAMELYON-16 and TCGA Lung Cancer datasets demonstrate that MHIM-MIL outperforms other latest methods in terms of performance and training cost. The code is available at: https://github.com/DearCaat/MHIM-MIL.
Authors: Fei Teng, Jiaming Zhang, Kunyu Peng, Yaonan Wang, Rainer Stiefelhagen, Kailun Yang
Light field cameras, by harnessing the power of micro-lens array, are capable of capturing intricate angular and spatial details. This allows for acquiring complex light patterns and details from multiple angles, significantly enhancing the precision of image semantic segmentation, a critical aspect of scene interpretation in vision intelligence. However, the extensive angular information of light field cameras contains a large amount of redundant data, which is overwhelming for the limited hardware resources of intelligent vehicles. Besides, inappropriate compression leads to information corruption and data loss. To excavate representative information, we propose a new paradigm, Omni-Aperture Fusion model (OAFuser), which leverages dense context from the central view and discovers the angular information from sub-aperture images to generate a semantically consistent result. To avoid feature loss during network propagation and simultaneously streamline the redundant information from the light field camera, we present a simple yet very effective Sub-Aperture Fusion Module (SAFM) to embed sub-aperture images into angular features without any additional memory cost. Furthermore, to address the mismatched spatial information across viewpoints, we present a Center Angular Rectification Module (CARM) to realize feature resorting and prevent feature occlusion caused by asymmetric information. Our proposed OAFuser achieves state-of-the-art performance on the UrbanLF-Real and -Syn datasets and sets a new record of 84.93% in mIoU on the UrbanLF-Real Extended dataset, with a gain of +4.53%. The source code of OAFuser will be available at https://github.com/FeiBryantkit/OAFuser.
Authors: Shili Zhou, Ruian He, Weimin Tan, Bo Yan
Optical Flow Estimation aims to find the 2D dense motion field between two frames. Due to the limitation of model structures and training datasets, existing methods often rely too much on local clues and ignore the integrity of objects, resulting in fragmented motion estimation. Through theoretical analysis, we find the pre-trained large vision models are helpful in optical flow estimation, and we notice that the recently famous Segment Anything Model (SAM) demonstrates a strong ability to segment complete objects, which is suitable for solving the fragmentation problem. We thus propose a solution to embed the frozen SAM image encoder into FlowFormer to enhance object perception. To address the challenge of in-depth utilizing SAM in non-segmentation tasks like optical flow estimation, we propose an Optical Flow Task-Specific Adaption scheme, including a Context Fusion Module to fuse the SAM encoder with the optical flow context encoder, and a Context Adaption Module to adapt the SAM features for optical flow task with Learned Task-Specific Embedding. Our proposed SAMFlow model reaches 0.86/2.10 clean/final EPE and 3.55/12.32 EPE/F1-all on Sintel and KITTI-15 training set, surpassing Flowformer by 8.5%/9.9% and 13.2%/16.3%. Furthermore, our model achieves state-of-the-art performance on the Sintel and KITTI-15 benchmarks, ranking #1 among all two-frame methods on Sintel clean pass.
Authors: Jorge Paz-Ruza, Amparo Alonso-Betanzos, Berta Guijarro-Berdiñas, Brais Cancela, Carlos Eiras-Franco
Recommender Systems have become crucial in the modern world, commonly guiding users towards relevant content or products, and having a large influence over the decisions of users and citizens. However, ensuring transparency and user trust in these systems remains a challenge; personalized explanations have emerged as a solution, offering justifications for recommendations. Among the existing approaches for generating personalized explanations, using existing visual content created by users is a promising option to maximize transparency and user trust. State-of-the-art models that follow this approach, despite leveraging highly optimized architectures, employ surrogate learning tasks that do not efficiently model the objective of ranking images as explanations for a given recommendation; this leads to a suboptimal training process with high computational costs that may not be reduced without affecting model performance. This work presents BRIE, a novel model where we leverage Bayesian Pairwise Ranking to enhance the training process, allowing us to consistently outperform state-of-the-art models in six real-world datasets while reducing its model size by up to 64 times and its CO${_2}$ emissions by up to 75% in training and inference.
Authors: Jiajia Li, Mingle Xu, Lirong Xiang, Dong Chen, Weichao Zhuang, Xunyuan Yin, Zhaojian Li
The past decade has witnessed the rapid development of ML and DL methodologies in agricultural systems, showcased by great successes in variety of agricultural applications. However, these conventional ML/DL models have certain limitations: They heavily rely on large, costly-to-acquire labeled datasets for training, require specialized expertise for development and maintenance, and are mostly tailored for specific tasks, thus lacking generalizability. Recently, foundation models have demonstrated remarkable successes in language and vision tasks across various domains. These models are trained on a vast amount of data from multiple domains and modalities. Once trained, they can accomplish versatile tasks with just minor fine-tuning and minimal task-specific labeled data. Despite their proven effectiveness and huge potential, there has been little exploration of applying FMs to agriculture fields. Therefore, this study aims to explore the potential of FMs in the field of smart agriculture. In particular, we present conceptual tools and technical background to facilitate the understanding of the problem space and uncover new research directions in this field. To this end, we first review recent FMs in the general computer science domain and categorize them into four categories: language FMs, vision FMs, multimodal FMs, and reinforcement learning FMs. Subsequently, we outline the process of developing agriculture FMs and discuss their potential applications in smart agriculture. We also discuss the unique challenges associated with developing AFMs, including model training, validation, and deployment. Through this study, we contribute to the advancement of AI in agriculture by introducing AFMs as a promising paradigm that can significantly mitigate the reliance on extensive labeled datasets and enhance the efficiency, effectiveness, and generalization of agricultural AI systems.
Authors: Andre Ye, Quan Ze Chen, Amy Zhang
Medical image segmentation modeling is a high-stakes task where understanding of uncertainty is crucial for addressing visual ambiguity. Prior work has developed segmentation models utilizing probabilistic or generative mechanisms to infer uncertainty from labels where annotators draw a singular boundary. However, as these annotations cannot represent an individual annotator's uncertainty, models trained on them produce uncertainty maps that are difficult to interpret. We propose a novel segmentation representation, Confidence Contours, which uses high- and low-confidence ``contours'' to capture uncertainty directly, and develop a novel annotation system for collecting contours. We conduct an evaluation on the Lung Image Dataset Consortium (LIDC) and a synthetic dataset. From an annotation study with 30 participants, results show that Confidence Contours provide high representative capacity without considerably higher annotator effort. We also find that general-purpose segmentation models can learn Confidence Contours at the same performance level as standard singular annotations. Finally, from interviews with 5 medical experts, we find that Confidence Contour maps are more interpretable than Bayesian maps due to representation of structural uncertainty.
Authors: Anubhav Jain, Nasir Memon, Julian Togelius
Deep generative models require large amounts of training data. This often poses a problem as the collection of datasets can be expensive and difficult, in particular datasets that are representative of the appropriate underlying distribution (e.g. demographic). This introduces biases in datasets which are further propagated in the models. We present an approach to construct an unbiased generative adversarial network (GAN) from an existing biased GAN by rebalancing the model distribution. We do so by generating balanced data from an existing imbalanced deep generative model using an evolutionary algorithm and then using this data to train a balanced generative model. Additionally, we propose a bias mitigation loss function that minimizes the deviation of the learned class distribution from being equiprobable. We show results for the StyleGAN2 models while training on the Flickr Faces High Quality (FFHQ) dataset for racial fairness and see that the proposed approach improves on the fairness metric by almost 5 times, whilst maintaining image quality. We further validate our approach by applying it to an imbalanced CIFAR10 dataset where we show that we can obtain comparable fairness and image quality as when training on a balanced CIFAR10 dataset which is also twice as large. Lastly, we argue that the traditionally used image quality metrics such as Frechet inception distance (FID) are unsuitable for scenarios where the class distributions are imbalanced and a balanced reference set is not available.
Authors: Wenxi Yue, Jing Zhang, Kun Hu, Yong Xia, Jiebo Luo, Zhiyong Wang
The Segment Anything Model (SAM) is a powerful foundation model that has revolutionised image segmentation. To apply SAM to surgical instrument segmentation, a common approach is to locate precise points or boxes of instruments and then use them as prompts for SAM in a zero-shot manner. However, we observe two problems with this naive pipeline: (1) the domain gap between natural objects and surgical instruments leads to inferior generalisation of SAM; and (2) SAM relies on precise point or box locations for accurate segmentation, requiring either extensive manual guidance or a well-performing specialist detector for prompt preparation, which leads to a complex multi-stage pipeline. To address these problems, we introduce SurgicalSAM, a novel end-to-end efficient-tuning approach for SAM to effectively integrate surgical-specific information with SAM's pre-trained knowledge for improved generalisation. Specifically, we propose a lightweight prototype-based class prompt encoder for tuning, which directly generates prompt embeddings from class prototypes and eliminates the use of explicit prompts for improved robustness and a simpler pipeline. In addition, to address the low inter-class variance among surgical instrument categories, we propose contrastive prototype learning, further enhancing the discrimination of the class prototypes for more accurate class prompting. The results of extensive experiments on both EndoVis2018 and EndoVis2017 datasets demonstrate that SurgicalSAM achieves state-of-the-art performance while only requiring a small number of tunable parameters. The source code is available at https://github.com/wenxi-yue/SurgicalSAM.
Authors: Min Cao, Yang Bai, Ziyin Zeng, Mang Ye, Min Zhang
Text-based Person Search (TBPS) aims to retrieve the person images using natural language descriptions. Recently, Contrastive Language Image Pretraining (CLIP), a universal large cross-modal vision-language pre-training model, has remarkably performed over various cross-modal downstream tasks due to its powerful cross-modal semantic learning capacity. TPBS, as a fine-grained cross-modal retrieval task, is also facing the rise of research on the CLIP-based TBPS. In order to explore the potential of the visual-language pre-training model for downstream TBPS tasks, this paper makes the first attempt to conduct a comprehensive empirical study of CLIP for TBPS and thus contribute a straightforward, incremental, yet strong TBPS-CLIP baseline to the TBPS community. We revisit critical design considerations under CLIP, including data augmentation and loss function. The model, with the aforementioned designs and practical training tricks, can attain satisfactory performance without any sophisticated modules. Also, we conduct the probing experiments of TBPS-CLIP in model generalization and model compression, demonstrating the effectiveness of TBPS-CLIP from various aspects. This work is expected to provide empirical insights and highlight future CLIP-based TBPS research.
Authors: Harrison Rosenberg, Shimaa Ahmed, Guruprasad V Ramesh, Ramya Korlakai Vinayak, Kassem Fawaz
Text-to-image diffusion models have achieved widespread popularity due to their unprecedented image generation capability. In particular, their ability to synthesize and modify human faces has spurred research into using generated face images in both training data augmentation and model performance assessments. In this paper, we study the efficacy and shortcomings of generative models in the context of face generation. Utilizing a combination of qualitative and quantitative measures, including embedding-based metrics and user studies, we present a framework to audit the characteristics of generated faces conditioned on a set of social attributes. We applied our framework on faces generated through state-of-the-art text-to-image diffusion models. We identify several limitations of face image generation that include faithfulness to the text prompt, demographic disparities, and distributional shifts. Furthermore, we present an analytical model that provides insights into how training data selection contributes to the performance of generative models.
Authors: Wenzhang Wei, Zhipeng Gui, Changguang Wu, Anqi Zhao, Dehua Peng, Huayi Wu
The core of cross-modal matching is to accurately measure the similarity between different modalities in a unified representation space. However, compared to textual descriptions of a certain perspective, the visual modality has more semantic variations. So, images are usually associated with multiple textual captions in databases. Although popular symmetric embedding methods have explored numerous modal interaction approaches, they often learn toward increasing the average expression probability of multiple semantic variations within image embeddings. Consequently, information entropy in embeddings is increased, resulting in redundancy and decreased accuracy. In this work, we propose a Dynamic Visual Semantic Sub-Embeddings framework (DVSE) to reduce the information entropy. Specifically, we obtain a set of heterogeneous visual sub-embeddings through dynamic orthogonal constraint loss. To encourage the generated candidate embeddings to capture various semantic variations, we construct a mixed distribution and employ a variance-aware weighting loss to assign different weights to the optimization process. In addition, we develop a Fast Re-ranking strategy (FR) to efficiently evaluate the retrieval results and enhance the performance. We compare the performance with existing set-based method using four image feature encoders and two text feature encoders on three benchmark datasets: MSCOCO, Flickr30K and CUB Captions. We also show the role of different components by ablation studies and perform a sensitivity analysis of the hyperparameters. The qualitative analysis of visualized bidirectional retrieval and attention maps further demonstrates the ability of our method to encode semantic variations.
Authors: Xingjian Diao, Ming Cheng, Shitong Cheng
Learning high-quality video representation has shown significant applications in computer vision and remains challenging. Previous work based on mask autoencoders such as ImageMAE and VideoMAE has proven the effectiveness of learning representations in images and videos through reconstruction strategy in the visual modality. However, these models exhibit inherent limitations, particularly in scenarios where extracting features solely from the visual modality proves challenging, such as when dealing with low-resolution and blurry original videos. Based on this, we propose AV-MaskEnhancer for learning high-quality video representation by combining visual and audio information. Our approach addresses the challenge by demonstrating the complementary nature of audio and video features in cross-modality content. Moreover, our result of the video classification task on the UCF101 dataset outperforms the existing work and reaches the state-of-the-art, with a top-1 accuracy of 98.8% and a top-5 accuracy of 99.9%.
Authors: Mengyue Yang, Zhen Fang, Yonggang Zhang, Yali Du, Furui Liu, Jean-Francois Ton, Jianhong Wang, Jun Wang
Out-of-distribution (OOD) generalization is indispensable for learning models in the wild, where testing distribution typically unknown and different from the training. Recent methods derived from causality have shown great potential in achieving OOD generalization. However, existing methods mainly focus on the invariance property of causes, while largely overlooking the property of \textit{sufficiency} and \textit{necessity} conditions. Namely, a necessary but insufficient cause (feature) is invariant to distribution shift, yet it may not have required accuracy. By contrast, a sufficient yet unnecessary cause (feature) tends to fit specific data well but may have a risk of adapting to a new domain. To capture the information of sufficient and necessary causes, we employ a classical concept, the probability of sufficiency and necessary causes (PNS), which indicates the probability of whether one is the necessary and sufficient cause. To associate PNS with OOD generalization, we propose PNS risk and formulate an algorithm to learn representation with a high PNS value. We theoretically analyze and prove the generalizability of the PNS risk. Experiments on both synthetic and real-world benchmarks demonstrate the effectiveness of the proposed method. The details of the implementation can be found at the GitHub repository: https://github.com/ymy4323460/CaSN.
Authors: Haoxuan Qu, Xiaofei Hui, Yujun Cai, Jun Liu
Open-set object recognition aims to identify if an object is from a class that has been encountered during training or not. To perform open-set object recognition accurately, a key challenge is how to reduce the reliance on spurious-discriminative features. In this paper, motivated by that different large models pre-trained through different paradigms can possess very rich while distinct implicit knowledge, we propose a novel framework named Large Model Collaboration (LMC) to tackle the above challenge via collaborating different off-the-shelf large models in a training-free manner. Moreover, we also incorporate the proposed framework with several novel designs to effectively extract implicit knowledge from large models. Extensive experiments demonstrate the efficacy of our proposed framework. Code is available https://github.com/Harryqu123/LMC
Authors: Vasantha Ramani, Pandarasamy Arjunan, Kameshwar Poolla, Clayton Miller
This work presents the analysis of semantically segmented, longitudinally, and spatially rich thermal images collected at the neighborhood scale to identify hot and cool spots in urban areas. An infrared observatory was operated over a few months to collect thermal images of different types of buildings on the educational campus of the National University of Singapore. A subset of the thermal image dataset was used to train state-of-the-art deep learning models to segment various urban features such as buildings, vegetation, sky, and roads. It was observed that the U-Net segmentation model with `resnet34' CNN backbone has the highest mIoU score of 0.99 on the test dataset, compared to other models such as DeepLabV3, DeeplabV3+, FPN, and PSPnet. The masks generated using the segmentation models were then used to extract the temperature from thermal images and correct for differences in the emissivity of various urban features. Further, various statistical measure of the temperature extracted using the predicted segmentation masks is shown to closely match the temperature extracted using the ground truth masks. Finally, the masks were used to identify hot and cool spots in the urban feature at various instances of time. This forms one of the very few studies demonstrating the automated analysis of thermal images, which can be of potential use to urban planners for devising mitigation strategies for reducing the urban heat island (UHI) effect, improving building energy efficiency, and maximizing outdoor thermal comfort.
Authors: Mehdi Fatan, Emanuele Mincato, Dimitra Pintzou, Mariella Dimiccoli
Predicting turn-taking in multiparty conversations has many practical applications in human-computer/robot interaction. However, the complexity of human communication makes it a challenging task. Recent advances have shown that synchronous multi-perspective egocentric data can significantly improve turn-taking prediction compared to asynchronous, single-perspective transcriptions. Building on this research, we propose a new multimodal transformer-based architecture for predicting turn-taking in embodied, synchronized multi-perspective data. Our experimental results on the recently introduced EgoCom dataset show a substantial performance improvement of up to 14.01% on average compared to existing baselines and alternative transformer-based approaches. The source code, and the pre-trained models of our 3M-Transformer will be available upon acceptance.
Authors: Yinan Liang, Ziwei Wang, Xiuwei Xu, Yansong Tang, Jie Zhou, Jiwen Lu
Due to the high price and heavy energy consumption of GPUs, deploying deep models on IoT devices such as microcontrollers makes significant contributions for ecological AI. Conventional methods successfully enable convolutional neural network inference of high resolution images on microcontrollers, while the framework for vision transformers that achieve the state-of-the-art performance in many vision applications still remains unexplored. In this paper, we propose a hardware-algorithm co-optimizations method called MCUFormer to deploy vision transformers on microcontrollers with extremely limited memory, where we jointly design transformer architecture and construct the inference operator library to fit the memory resource constraint. More specifically, we generalize the one-shot network architecture search (NAS) to discover the optimal architecture with highest task performance given the memory budget from the microcontrollers, where we enlarge the existing search space of vision transformers by considering the low-rank decomposition dimensions and patch resolution for memory reduction. For the construction of the inference operator library of vision transformers, we schedule the memory buffer during inference through operator integration, patch embedding decomposition, and token overwriting, allowing the memory buffer to be fully utilized to adapt to the forward pass of the vision transformer. Experimental results demonstrate that our MCUFormer achieves 73.62\% top-1 accuracy on ImageNet for image classification with 320KB memory on STM32F746 microcontroller. Code is available at https://github.com/liangyn22/MCUFormer.
Authors: Vibhas K. Vats, Sripad Joshi, David J. Crandall, Md. Alimoor Reza, Soon-heung Jung
Traditional multi-view stereo (MVS) methods rely heavily on photometric and geometric consistency constraints, but newer machine learning-based MVS methods check geometric consistency across multiple source views only as a post-processing step. In this paper, we present a novel approach that explicitly encourages geometric consistency of reference view depth maps across multiple source views at different scales during learning (see Fig. 1). We find that adding this geometric consistency loss significantly accelerates learning by explicitly penalizing geometrically inconsistent pixels, reducing the training iteration requirements to nearly half that of other MVS methods. Our extensive experiments show that our approach achieves a new state-of-the-art on the DTU and BlendedMVS datasets, and competitive results on the Tanks and Temples benchmark. To the best of our knowledge, GC-MVSNet is the first attempt to enforce multi-view, multi-scale geometric consistency during learning.
Authors: Chun-Chuen Hui, Wan-Chi Siu, Ngai-Fong Law
Large scale image super-resolution is a challenging computer vision task, since vast information is missing in a highly degraded image, say for example forscale x16 super-resolution. Diffusion models are used successfully in recent years in extreme super-resolution applications, in which Gaussian noise is used as a means to form a latent photo-realistic space, and acts as a link between the space of latent vectors and the latent photo-realistic space. There are quite a few sophisticated mathematical derivations on mapping the statistics of Gaussian noises making Diffusion Models successful. In this paper we propose a simple approach which gets away from using Gaussian noise but adopts some basic structures of diffusion models for efficient image super-resolution. Essentially, we propose a DNN to perform domain transfer between neighbor domains, which can learn the differences in statistical properties to facilitate gradual interpolation with results of reasonable quality. Further quality improvement is achieved by conditioning the domain transfer with reference to the input LR image. Experimental results show that our method outperforms not only state-of-the-art large scale super resolution models, but also the current diffusion models for image super-resolution. The approach can readily be extended to other image-to-image tasks, such as image enlightening, inpainting, denoising, etc.
Authors: Shengzhe Zhou, Zejian Lee, Shengyuan Zhang, Lefan Hou, Changyuan Yang, Guang Yang, Zhiyuan Yang, Lingyun Sun
Denoising Diffusion models have exhibited remarkable capabilities in image generation. However, generating high-quality samples requires a large number of iterations. Knowledge distillation for diffusion models is an effective method to address this limitation with a shortened sampling process but causes degraded generative quality. Based on our analysis with bias-variance decomposition and experimental observations, we attribute the degradation to the spatial fitting error occurring in the training of both the teacher and student model. Accordingly, we propose $\textbf{S}$patial $\textbf{F}$itting-$\textbf{E}$rror $\textbf{R}$eduction $\textbf{D}$istillation model ($\textbf{SFERD}$). SFERD utilizes attention guidance from the teacher model and a designed semantic gradient predictor to reduce the student's fitting error. Empirically, our proposed model facilitates high-quality sample generation in a few function evaluations. We achieve an FID of 5.31 on CIFAR-10 and 9.39 on ImageNet 64$\times$64 with only one step, outperforming existing diffusion methods. Our study provides a new perspective on diffusion distillation by highlighting the intrinsic denoising ability of models. Project link: \url{https://github.com/Sainzerjj/SFERD}.
Authors: Haoyi Duan, Yan Xia, Mingze Zhou, Li Tang, Jieming Zhu, Zhou Zhao
In recent years, the deployment of large-scale pre-trained models in audio-visual downstream tasks has yielded remarkable outcomes. However, these models, primarily trained on single-modality unconstrained datasets, still encounter challenges in feature extraction for multi-modal tasks, leading to suboptimal performance. This limitation arises due to the introduction of irrelevant modality-specific information during encoding, which adversely affects the performance of downstream tasks. To address this challenge, this paper proposes a novel Dual-Guided Spatial-Channel-Temporal (DG-SCT) attention mechanism. This mechanism leverages audio and visual modalities as soft prompts to dynamically adjust the parameters of pre-trained models based on the current multi-modal input features. Specifically, the DG-SCT module incorporates trainable cross-modal interaction layers into pre-trained audio-visual encoders, allowing adaptive extraction of crucial information from the current modality across spatial, channel, and temporal dimensions, while preserving the frozen parameters of large-scale pre-trained models. Experimental evaluations demonstrate that our proposed model achieves state-of-the-art results across multiple downstream tasks, including AVE, AVVP, AVS, and AVQA. Furthermore, our model exhibits promising performance in challenging few-shot and zero-shot scenarios. The source code and pre-trained models are available at https://github.com/haoyi-duan/DG-SCT.
Authors: Martin Cubaud (LaSTIG), Arnaud Le Bris (LaSTIG), Laurence Jolivet (LaSTIG), Ana-Maria Olteanu-Raimond (LaSTIG)
Accurate land use maps, describing the territory from an anthropic utilisation point of view, are useful tools for land management and planning. To produce them, the use of optical images alone remains limited. It is therefore necessary to make use of several heterogeneous sources, each carrying complementary or contradictory information due to their imperfections or their different specifications. This study compares two different approaches i.e. a pre-classification and a post-classification fusion approach for combining several sources of spatial data in the context of land use classification. The approaches are applied on authoritative land use data located in the Gers department in the southwest of France. Pre-classification fusion, while not explicitly modeling imperfections, has the best final results, reaching an overall accuracy of 97% and a macro-mean F1 score of 88%.
Authors: Shreshth Saini, Avinab Saha, Alan C. Bovik
We introduce HIDRO-VQA, a no-reference (NR) video quality assessment model designed to provide precise quality evaluations of High Dynamic Range (HDR) videos. HDR videos exhibit a broader spectrum of luminance, detail, and color than Standard Dynamic Range (SDR) videos. As HDR content becomes increasingly popular, there is a growing demand for video quality assessment (VQA) algorithms that effectively address distortions unique to HDR content. To address this challenge, we propose a self-supervised contrastive fine-tuning approach to transfer quality-aware features from the SDR to the HDR domain, utilizing unlabeled HDR videos. Our findings demonstrate that self-supervised pre-trained neural networks on SDR content can be further fine-tuned in a self-supervised setting using limited unlabeled HDR videos to achieve state-of-the-art performance on the only publicly available VQA database for HDR content, the LIVE-HDR VQA database. Moreover, our algorithm can be extended to the Full Reference VQA setting, also achieving state-of-the-art performance. Our code is available publicly at https://github.com/avinabsaha/HIDRO-VQA.
Authors: Luan Wei, Anna Hilsmann, Peter Eisert
Piece-wise planar 3D reconstruction simultaneously segments plane instances and recovers their 3D plane parameters from an image, which is particularly useful for indoor or man-made environments. Efficient reconstruction of 3D planes coupled with semantic predictions offers advantages for a wide range of applications requiring scene understanding and concurrent spatial mapping. However, most existing planar reconstruction models either neglect semantic predictions or do not run efficiently enough for real-time applications. We introduce SOLOPlanes, a real-time planar reconstruction model based on a modified instance segmentation architecture which simultaneously predicts semantics for each plane instance, along with plane parameters and piece-wise plane instance masks. We achieve an improvement in instance mask segmentation by including multi-view guidance for plane predictions in the training process. This cross-task improvement, training for plane prediction but improving the mask segmentation, is due to the nature of feature sharing in multi-task learning. Our model simultaneously predicts semantics using single images at inference time, while achieving real-time predictions at 43 FPS.
Authors: Haoze Sun, Wenbo Li, Jianzhuang Liu, Haoyu Chen, Renjing Pei, Xueyi Zou, Youliang Yan, Yujiu Yang
Existing super-resolution (SR) models primarily focus on restoring local texture details, often neglecting the global semantic information within the scene. This oversight can lead to the omission of crucial semantic details or the introduction of inaccurate textures during the recovery process. In our work, we introduce the Cognitive Super-Resolution (CoSeR) framework, empowering SR models with the capacity to comprehend low-resolution images. We achieve this by marrying image appearance and language understanding to generate a cognitive embedding, which not only activates prior information from large text-to-image diffusion models but also facilitates the generation of high-quality reference images to optimize the SR process. To further improve image fidelity, we propose a novel condition injection scheme called "All-in-Attention", consolidating all conditional information into a single module. Consequently, our method successfully restores semantically correct and photorealistic details, demonstrating state-of-the-art performance across multiple benchmarks. Code: https://github.com/VINHYU/CoSeR
Authors: Ryutaro Tanno, David G.T. Barrett, Andrew Sellergren, Sumedh Ghaisas, Sumanth Dathathri, Abigail See, Johannes Welbl, Karan Singhal, Shekoofeh Azizi, Tao Tu, Mike Schaekermann, Rhys May, Roy Lee, SiWai Man, Zahra Ahmed, Sara Mahdavi, Yossi Matias, Joelle Barral, Ali Eslami, Danielle Belgrave, Vivek Natarajan, Shravya Shetty, Pushmeet Kohli, Po-Sen Huang, Alan Karthikesalingam, Ira Ktena
Radiology reports are an instrumental part of modern medicine, informing key clinical decisions such as diagnosis and treatment. The worldwide shortage of radiologists, however, restricts access to expert care and imposes heavy workloads, contributing to avoidable errors and delays in report delivery. While recent progress in automated report generation with vision-language models offer clear potential in ameliorating the situation, the path to real-world adoption has been stymied by the challenge of evaluating the clinical quality of AI-generated reports. In this study, we build a state-of-the-art report generation system for chest radiographs, $\textit{Flamingo-CXR}$, by fine-tuning a well-known vision-language foundation model on radiology data. To evaluate the quality of the AI-generated reports, a group of 16 certified radiologists provide detailed evaluations of AI-generated and human written reports for chest X-rays from an intensive care setting in the United States and an inpatient setting in India. At least one radiologist (out of two per case) preferred the AI report to the ground truth report in over 60$\%$ of cases for both datasets. Amongst the subset of AI-generated reports that contain errors, the most frequently cited reasons were related to the location and finding, whereas for human written reports, most mistakes were related to severity and finding. This disparity suggested potential complementarity between our AI system and human experts, prompting us to develop an assistive scenario in which Flamingo-CXR generates a first-draft report, which is subsequently revised by a clinician. This is the first demonstration of clinician-AI collaboration for report writing, and the resultant reports are assessed to be equivalent or preferred by at least one radiologist to reports written by experts alone in 80$\%$ of in-patient cases and 60$\%$ of intensive care cases.
Authors: Yushi Lan, Feitong Tan, Di Qiu, Qiangeng Xu, Kyle Genova, Zeng Huang, Sean Fanello, Rohit Pandey, Thomas Funkhouser, Chen Change Loy, Yinda Zhang
We present a novel framework for generating photorealistic 3D human head and subsequently manipulating and reposing them with remarkable flexibility. The proposed approach leverages an implicit function representation of 3D human heads, employing 3D Gaussians anchored on a parametric face model. To enhance representational capabilities and encode spatial information, we embed a lightweight tri-plane payload within each Gaussian rather than directly storing color and opacity. Additionally, we parameterize the Gaussians in a 2D UV space via a 3DMM, enabling effective utilization of the diffusion model for 3D head avatar generation. Our method facilitates the creation of diverse and realistic 3D human heads with fine-grained editing over facial features and expressions. Extensive experiments demonstrate the effectiveness of our method.
Authors: Linze Li, Sunqi Fan, Hengjun Pu, Zhaodong Bing, Yao Tang, Tianzhu Ye, Tong Yang, Liangyu Chen, Jiajun Liang
Over recent years, diffusion models have facilitated significant advancements in video generation. Yet, the creation of face-related videos still confronts issues such as low facial fidelity, lack of frame consistency, limited editability and uncontrollable human poses. To address these challenges, we introduce a facial animation generation method that enhances both face identity fidelity and editing capabilities while ensuring frame consistency. This approach incorporates the concept of an anchor frame to counteract the degradation of generative ability in original text-to-image models when incorporating a motion module. We propose two strategies towards this objective: training-free and training-based anchor frame methods. Our method's efficacy has been validated on multiple representative DreamBooth and LoRA models, delivering substantial improvements over the original outcomes in terms of facial fidelity, text-to-image editability, and video motion. Moreover, we introduce conditional control using a 3D parametric face model to capture accurate facial movements and expressions. This solution augments the creative possibilities for facial animation generation through the integration of multiple control signals. For additional samples, please visit https://paper-faac.github.io/.
Authors: Qirui Yang, Qihua Cheng, Huanjing Yue, Le Zhang, Yihao Liu, Jingyu Yang
Raw low light image enhancement (LLIE) has achieved much better performance than the sRGB domain enhancement methods due to the merits of raw data. However, the ambiguity between noisy to clean and raw to sRGB mappings may mislead the single-stage enhancement networks. The two-stage networks avoid ambiguity by decoupling the two mappings but usually have large computing complexity. To solve this problem, we propose a single-stage network empowered by Feature Domain Adaptation (FDA) to decouple the denoising and color mapping tasks in raw LLIE. The denoising encoder is supervised by the clean raw image, and then the denoised features are adapted for the color mapping task by an FDA module. We propose a Lineformer to serve as the FDA, which can well explore the global and local correlations with fewer line buffers (friendly to the line-based imaging process). During inference, the raw supervision branch is removed. In this way, our network combines the advantage of a two-stage enhancement process with the efficiency of single-stage inference. Experiments on four benchmark datasets demonstrate that our method achieves state-of-the-art performance with fewer computing costs (60% FLOPs of the two-stage method DNF). Our codes will be released after the acceptance of this work.
Authors: Hao Shao, Yuxuan Hu, Letian Wang, Steven L. Waslander, Yu Liu, Hongsheng Li
Despite significant recent progress in the field of autonomous driving, modern methods still struggle and can incur serious accidents when encountering long-tail unforeseen events and challenging urban scenarios. On the one hand, large language models (LLM) have shown impressive reasoning capabilities that approach "Artificial General Intelligence". On the other hand, previous autonomous driving methods tend to rely on limited-format inputs (e.g. sensor data and navigation waypoints), restricting the vehicle's ability to understand language information and interact with humans. To this end, this paper introduces LMDrive, a novel language-guided, end-to-end, closed-loop autonomous driving framework. LMDrive uniquely processes and integrates multi-modal sensor data with natural language instructions, enabling interaction with humans and navigation software in realistic instructional settings. To facilitate further research in language-based closed-loop autonomous driving, we also publicly release the corresponding dataset which includes approximately 64K instruction-following data clips, and the LangAuto benchmark that tests the system's ability to handle complex instructions and challenging driving scenarios. Extensive closed-loop experiments are conducted to demonstrate LMDrive's effectiveness. To the best of our knowledge, we're the very first work to leverage LLMs for closed-loop end-to-end autonomous driving. Codes, models, and datasets can be found at https://github.com/opendilab/LMDrive
Authors: Yanzuo Lu, Meng Shen, Andy J Ma, Xiaohua Xie, Jian-Huang Lai
Universal domain adaptation (UniDA) is a practical but challenging problem, in which information about the relation between the source and the target domains is not given for knowledge transfer. Existing UniDA methods may suffer from the problems of overlooking intra-domain variations in the target domain and difficulty in separating between the similar known and unknown class. To address these issues, we propose a novel Mutual Learning Network (MLNet) with neighborhood invariance for UniDA. In our method, confidence-guided invariant feature learning with self-adaptive neighbor selection is designed to reduce the intra-domain variations for more generalizable feature representation. By using the cross-domain mixup scheme for better unknown-class identification, the proposed method compensates for the misidentified known-class errors by mutual learning between the closed-set and open-set classifiers. Extensive experiments on three publicly available benchmarks demonstrate that our method achieves the best results compared to the state-of-the-arts in most cases and significantly outperforms the baseline across all the four settings in UniDA. Code is available at https://github.com/YanzuoLu/MLNet.
Authors: Weilong Chai, DanDan Zheng, Jiajiong Cao, Zhiquan Chen, Changbao Wang, Chenguang Ma
Text-to-image diffusion models (SD) exhibit significant advancements while requiring extensive computational resources. Though many acceleration methods have been proposed, they suffer from generation quality degradation or extra training cost generalizing to new fine-tuned models. To address these limitations, we propose a novel and universal Stable-Diffusion (SD) acceleration module called SpeedUpNet(SUN). SUN can be directly plugged into various fine-tuned SD models without extra training. This technique utilizes cross-attention layers to learn the relative offsets in the generated image results between negative and positive prompts achieving classifier-free guidance distillation with negative prompts controllable, and introduces a Multi-Step Consistency (MSC) loss to ensure a harmonious balance between reducing inference steps and maintaining consistency in the generated output. Consequently, SUN significantly reduces the number of inference steps to just 4 steps and eliminates the need for classifier-free guidance. It leads to an overall speedup of more than 10 times for SD models compared to the state-of-the-art 25-step DPM-solver++, and offers two extra advantages: (1) classifier-free guidance distillation with controllable negative prompts and (2) seamless integration into various fine-tuned Stable-Diffusion models without training. The effectiveness of the SUN has been verified through extensive experimentation. Project Page: https://williechai.github.io/speedup-plugin-for-stable-diffusions.github.io
Authors: Jingcai Guo, Qihua Zhou, Ruibing Li, Xiaocheng Lu, Ziming Liu, Junyang Chen, Xin Xie, Jie Zhang
This paper provides a novel parsimonious yet efficient design for zero-shot learning (ZSL), dubbed ParsNets, where we are interested in learning a composition of on-device friendly linear networks, each with orthogonality and low-rankness properties, to achieve equivalent or even better performance against existing deep models. Concretely, we first refactor the core module of ZSL, i.e., visual-semantics mapping function, into several base linear networks that correspond to diverse components of the semantic space, where the complex nonlinearity can be collapsed into simple local linearities. Then, to facilitate the generalization of local linearities, we construct a maximal margin geometry on the learned features by enforcing low-rank constraints on intra-class samples and high-rank constraints on inter-class samples, resulting in orthogonal subspaces for different classes and each subspace lies on a compact manifold. To enhance the model's adaptability and counterbalance over/under-fittings in ZSL, a set of sample-wise indicators is employed to select a sparse subset from these base linear networks to form a composite semantic predictor for each sample. Notably, maximal margin geometry can guarantee the diversity of features, and meanwhile, local linearities guarantee efficiency. Thus, our ParsNets can generalize better to unseen classes and can be deployed flexibly on resource-constrained devices. Theoretical explanations and extensive experiments are conducted to verify the effectiveness of the proposed method.
Authors: Arefeh Rezaei, Mohammad Javad Ahmadi, Amir Molaei, Hamid. D. Taghirad
This paper aims to present a novel pipeline for automated surgical skill assessment using video data and to showcase the effectiveness of the proposed approach in evaluating surgeon proficiency, its potential for targeted training interventions, and quality assurance in surgical departments. The pipeline incorporates a representation flow convolutional neural network and a novel tree-based Gaussian process classifier, which is robust to noise, while being computationally efficient. Additionally, new kernels are introduced to enhance accuracy. The performance of the pipeline is evaluated using the JIGSAWS dataset. Comparative analysis with existing literature reveals significant improvement in accuracy and betterment in computation cost. The proposed pipeline contributes to computational efficiency and accuracy improvement in surgical skill assessment using video data. Results of our study based on comments of our colleague surgeons show that the proposed method has the potential to facilitate skill improvement among surgery fellows and enhance patient safety through targeted training interventions and quality assurance in surgical departments.
Authors: Yixin Zhang, Shen Zhao, Hanxue Gu, Maciej A. Mazurowski
Deep neural networks (DNNs) have been deployed for many image segmentation tasks and achieved outstanding performance. However, preparing a dataset for training segmentation DNNs is laborious and costly since typically pixel-level annotations are provided for each object of interest. To alleviate this issue, one can provide only weak labels such as bounding boxes or scribbles, or less accurate (noisy) annotations of the objects. These are significantly faster to generate and thus result in more annotated images given the same time budget. However, the reduction in quality might negatively affect the segmentation performance of the resulting model. In this study, we perform a thorough cost-effectiveness evaluation of several weak and noisy labels. We considered 11 variants of annotation strategies and 4 datasets. We conclude that the common practice of accurately outlining the objects of interest is virtually never the optimal approach when the annotation time is limited, even if notable annotation time is available (10s of hours). Annotation approaches that stood out in such scenarios were (1) contour-based annotation with rough continuous traces, (2) polygon-based annotation with few vertices, and (3) box annotations combined with the Segment Anything Model (SAM). In situations where unlimited annotation time was available, precise annotations still lead to the highest segmentation model performance.
Authors: Qi Mao, Lan Chen, Yuchao Gu, Zhen Fang, Mike Zheng Shou
Recent diffusion-based image editing approaches have exhibited impressive editing capabilities in images with simple compositions. However, localized editing in complex scenarios has not been well-studied in the literature, despite its growing real-world demands. Existing mask-based inpainting methods fall short of retaining the underlying structure within the edit region. Meanwhile, mask-free attention-based methods often exhibit editing leakage and misalignment in more complex compositions. In this work, we develop MAG-Edit, a training-free, inference-stage optimization method, which enables localized image editing in complex scenarios. In particular, MAG-Edit optimizes the noise latent feature in diffusion models by maximizing two mask-based cross-attention constraints of the edit token, which in turn gradually enhances the local alignment with the desired prompt. Extensive quantitative and qualitative experiments demonstrate the effectiveness of our method in achieving both text alignment and structure preservation for localized editing within complex scenarios.
Authors: Junfeng Long, Zirui Wang, Quanyi Li, Jiawei Gao, Liu Cao, Jiangmiao Pang
Robust locomotion control depends on accurate state estimations. However, the sensors of most legged robots can only provide partial and noisy observations, making the estimation particularly challenging, especially for external states like terrain frictions and elevation maps. Inspired by the classical Internal Model Control principle, we consider these external states as disturbances and introduce Hybrid Internal Model (HIM) to estimate them according to the response of the robot. The response, which we refer to as the hybrid internal embedding, contains the robot's explicit velocity and implicit stability representation, corresponding to two primary goals for locomotion tasks: explicitly tracking velocity and implicitly maintaining stability. We use contrastive learning to optimize the embedding to be close to the robot's successor state, in which the response is naturally embedded. HIM has several appealing benefits: It only needs the robot's proprioceptions, i.e., those from joint encoders and IMU as observations. It innovatively maintains consistent observations between simulation reference and reality that avoids information loss in mimicking learning. It exploits batch-level information that is more robust to noises and keeps better sample efficiency. It only requires 1 hour of training on an RTX 4090 to enable a quadruped robot to traverse any terrain under any disturbances. A wealth of real-world experiments demonstrates its agility, even in high-difficulty tasks and cases never occurred during the training process, revealing remarkable open-world generalizability.
Authors: Jiankai Sun, Chuanyang Zheng, Enze Xie, Zhengying Liu, Ruihang Chu, Jianing Qiu, Jiaqi Xu, Mingyu Ding, Hongyang Li, Mengzhe Geng, Yue Wu, Wenhai Wang, Junsong Chen, Zhangyue Yin, Xiaozhe Ren, Jie Fu, Junxian He, Wu Yuan, Qi Liu, Xihui Liu, Yu Li, Hao Dong, Yu Cheng, Ming Zhang, Pheng Ann Heng, Jifeng Dai, Ping Luo, Jingdong Wang, Ji-Rong Wen, Xipeng Qiu, Yike Guo, Hui Xiong, Qun Liu, Zhenguo Li
Reasoning, a crucial ability for complex problem-solving, plays a pivotal role in various real-world settings such as negotiation, medical diagnosis, and criminal investigation. It serves as a fundamental methodology in the field of Artificial General Intelligence (AGI). With the ongoing development of foundation models, there is a growing interest in exploring their abilities in reasoning tasks. In this paper, we introduce seminal foundation models proposed or adaptable for reasoning, highlighting the latest advancements in various reasoning tasks, methods, and benchmarks. We then delve into the potential future directions behind the emergence of reasoning abilities within foundation models. We also discuss the relevance of multimodal learning, autonomous agents, and super alignment in the context of reasoning. By discussing these future research directions, we hope to inspire researchers in their exploration of this field, stimulate further advancements in reasoning with foundation models, and contribute to the development of AGI.
Authors: David Charatan, Sizhe Li, Andrea Tagliasacchi, Vincent Sitzmann
We introduce pixelSplat, a feed-forward model that learns to reconstruct 3D radiance fields parameterized by 3D Gaussian primitives from pairs of images. Our model features real-time and memory-efficient rendering for scalable training as well as fast 3D reconstruction at inference time. To overcome local minima inherent to sparse and locally supported representations, we predict a dense probability distribution over 3D and sample Gaussian means from that probability distribution. We make this sampling operation differentiable via a reparameterization trick, allowing us to back-propagate gradients through the Gaussian splatting representation. We benchmark our method on wide-baseline novel view synthesis on the real-world RealEstate10k and ACID datasets, where we outperform state-of-the-art light field transformers and accelerate rendering by 2.5 orders of magnitude while reconstructing an interpretable and editable 3D radiance field.
Authors: Shutong Jin, Ruiyu Wang, Florian T. Pokorny
Although large-scale text-to-image generative models have shown promising performance in synthesizing high-quality images, directly applying these models to image editing remains a significant challenge. This challenge is further amplified in video editing due to the additional dimension of time. Especially for editing real videos as it necessitates maintaining a stable semantic layout across the frames while executing localized edits precisely without disrupting the existing backgrounds. In this paper, we propose RealCraft, an attention-control-based method for zero-shot editing in real videos. By employing the object-centric manipulation of cross-attention between prompts and frames and spatial-temporal attention within the frames, we achieve precise shape-wise editing along with enhanced consistency. Our model can be used directly with Stable Diffusion and operates without the need for additional localized information. We showcase our zero-shot attention-control-based method across a range of videos, demonstrating localized, high-fidelity, shape-precise and time-consistent editing in videos of various lengths, up to 64 frames.
Authors: Beibei Jing, Youjia Zhang, Zikai Song, Junqing Yu, Wei Yang
Generating realistic human motion sequences from text descriptions is a challenging task that requires capturing the rich expressiveness of both natural language and human motion.Recent advances in diffusion models have enabled significant progress in human motion synthesis.However, existing methods struggle to handle text inputs that describe complex or long motions.In this paper, we propose the Adaptable Motion Diffusion (AMD) model, which leverages a Large Language Model (LLM) to parse the input text into a sequence of concise and interpretable anatomical scripts that correspond to the target motion.This process exploits the LLM's ability to provide anatomical guidance for complex motion synthesis.We then devise a two-branch fusion scheme that balances the influence of the input text and the anatomical scripts on the inverse diffusion process, which adaptively ensures the semantic fidelity and diversity of the synthesized motion.Our method can effectively handle texts with complex or long motion descriptions, where existing methods often fail. Experiments on datasets with relatively more complex motions, such as CLCD1 and CLCD2, demonstrate that our AMD significantly outperforms existing state-of-the-art models.
Authors: Yuming Gu, Hongyi Xu, You Xie, Guoxian Song, Yichun Shi, Di Chang, Jing Yang, Linjie Luo
We present DiffPortrait3D, a conditional diffusion model that is capable of synthesizing 3D-consistent photo-realistic novel views from as few as a single in-the-wild portrait. Specifically, given a single RGB input, we aim to synthesize plausible but consistent facial details rendered from novel camera views with retained both identity and facial expression. In lieu of time-consuming optimization and fine-tuning, our zero-shot method generalizes well to arbitrary face portraits with unposed camera views, extreme facial expressions, and diverse artistic depictions. At its core, we leverage the generative prior of 2D diffusion models pre-trained on large-scale image datasets as our rendering backbone, while the denoising is guided with disentangled attentive control of appearance and camera pose. To achieve this, we first inject the appearance context from the reference image into the self-attention layers of the frozen UNets. The rendering view is then manipulated with a novel conditional control module that interprets the camera pose by watching a condition image of a crossed subject from the same view. Furthermore, we insert a trainable cross-view attention module to enhance view consistency, which is further strengthened with a novel 3D-aware noise generation process during inference. We demonstrate state-of-the-art results both qualitatively and quantitatively on our challenging in-the-wild and multi-view benchmarks.
Authors: Hongtao Wu, Ya Jing, Chilam Cheang, Guangzeng Chen, Jiafeng Xu, Xinghang Li, Minghuan Liu, Hang Li, Tao Kong
Generative pre-trained models have demonstrated remarkable effectiveness in language and vision domains by learning useful representations. In this paper, we extend the scope of this effectiveness by showing that visual robot manipulation can significantly benefit from large-scale video generative pre-training. We introduce GR-1, a straightforward GPT-style model designed for multi-task language-conditioned visual robot manipulation. GR-1 takes as inputs a language instruction, a sequence of observation images, and a sequence of robot states. It predicts robot actions as well as future images in an end-to-end manner. Thanks to a flexible design, GR-1 can be seamlessly finetuned on robot data after pre-trained on a large-scale video dataset. We perform extensive experiments on the challenging CALVIN benchmark and a real robot. On CALVIN benchmark, our method outperforms state-of-the-art baseline methods and improves the success rate from 88.9% to 94.9%. In the setting of zero-shot unseen scene generalization, GR-1 improves the success rate from 53.3% to 85.4%. In real robot experiments, GR-1 also outperforms baseline methods and shows strong potentials in generalization to unseen scenes and objects. We provide inaugural evidence that a unified GPT-style transformer, augmented with large-scale video generative pre-training, exhibits remarkable generalization to multi-task visual robot manipulation. Project page: https://GR1-Manipulation.github.io
Authors: Junwu Zhang, Zhenyu Tang, Yatian Pang, Xinhua Cheng, Peng Jin, Yida Wei, Munan Ning, Li Yuan
Recent one image to 3D generation methods commonly adopt Score Distillation Sampling (SDS). Despite the impressive results, there are multiple deficiencies including multi-view inconsistency, over-saturated and over-smoothed textures, as well as the slow generation speed. To address these deficiencies, we present Repaint123 to alleviate multi-view bias as well as texture degradation and speed up the generation process. The core idea is to combine the powerful image generation capability of the 2D diffusion model and the texture alignment ability of the repainting strategy for generating high-quality multi-view images with consistency. We further propose visibility-aware adaptive repainting strength for overlap regions to enhance the generated image quality in the repainting process. The generated high-quality and multi-view consistent images enable the use of simple Mean Square Error (MSE) loss for fast 3D content generation. We conduct extensive experiments and show that our method has a superior ability to generate high-quality 3D content with multi-view consistency and fine textures in 2 minutes from scratch. Our webpage is available at https://junwuzhang19.github.io/repaint123/.