Authors: Cuong V. Nguyen, Hieu Minh Duong, Cuong D.Do
We introduce MELEP, which stands for Muti-label Expected Log of Empirical Predictions, a novel measure to estimate how effective it is to transfer knowledge from a pre-trained model to a downstream task in a multi-label settings. The measure is generic to work with new target data having a different label set from source data. It is also computationally efficient, only requires forward passing the downstream dataset through the pre-trained model once. To the best of our knowledge, we are the first to develop such a transferability metric for multi-label ECG classification problems. Our experiments show that MELEP can predict the performance of pre-trained convolutional and recurrent deep neural networks, on small and imbalanced ECG data. Specifically, strong correlation coefficients, with absolute values exceeding 0.6 in most cases, were observed between MELEP and the actual average F1 scores of the fine-tuned models.
Authors: Yong Li, Zhiguo Zhao, Yunli Chen, Rui Tian
Spatial synchronization in roadside scenarios is essential for integrating data from multiple sensors at different locations. Current methods using cascading spatial transformation (CST) often lead to cumulative errors in large-scale deployments. Manual camera calibration is insufficient and requires extensive manual work, and existing methods are limited to controlled or single-view scenarios. To address these challenges, our research introduces a parallel spatial transformation (PST)-based framework for large-scale, multi-view, multi-sensor scenarios. PST parallelizes sensor coordinate system transformation, reducing cumulative errors. We incorporate deep learning for precise roadside monocular global localization, reducing manual work. Additionally, we use geolocation cues and an optimization algorithm for improved synchronization accuracy. Our framework has been tested in real-world scenarios, outperforming CST-based methods. It significantly enhances large-scale roadside multi-perspective, multi-sensor spatial synchronization, reducing deployment costs.
Authors: Yamin Li, Ange Lou, Catie Chang
In modern neuroscience, functional magnetic resonance imaging (fMRI) has been a crucial and irreplaceable tool that provides a non-invasive window into the dynamics of whole-brain activity. Nevertheless, fMRI is limited by hemodynamic blurring as well as high cost, immobility, and incompatibility with metal implants. Electroencephalography (EEG) is complementary to fMRI and can directly record the cortical electrical activity at high temporal resolution, but has more limited spatial resolution and is unable to recover information about deep subcortical brain structures. The ability to obtain fMRI information from EEG would enable cost-effective, imaging across a wider set of brain regions. Further, beyond augmenting the capabilities of EEG, cross-modality models would facilitate the interpretation of fMRI signals. However, as both EEG and fMRI are high-dimensional and prone to artifacts, it is currently challenging to model fMRI from EEG. To address this challenge, we propose a novel architecture that can predict fMRI signals directly from multi-channel EEG without explicit feature engineering. Our model achieves this by implementing a Sinusoidal Representation Network (SIREN) to learn frequency information in brain dynamics from EEG, which serves as the input to a subsequent encoder-decoder to effectively reconstruct the fMRI signal from a specific brain region. We evaluate our model using a simultaneous EEG-fMRI dataset with 8 subjects and investigate its potential for predicting subcortical fMRI signals. The present results reveal that our model outperforms a recent state-of-the-art model, and indicates the potential of leveraging periodic activation functions in deep neural networks to model functional neuroimaging data.
Authors: Khiem Vuong, Robert Tamburo, Srinivasa G. Narasimhan
Despite the widespread deployment of outdoor cameras, their potential for automated analysis remains largely untapped due, in part, to calibration challenges. The absence of precise camera calibration data, including intrinsic and extrinsic parameters, hinders accurate real-world distance measurements from captured videos. To address this, we present a scalable framework that utilizes street-level imagery to reconstruct a metric 3D model, facilitating precise calibration of in-the-wild traffic cameras. Notably, our framework achieves 3D scene reconstruction and accurate localization of over 100 global traffic cameras and is scalable to any camera with sufficient street-level imagery. For evaluation, we introduce a dataset of 20 fully calibrated traffic cameras, demonstrating our method's significant enhancements over existing automatic calibration techniques. Furthermore, we highlight our approach's utility in traffic analysis by extracting insights via 3D vehicle reconstruction and speed measurement, thereby opening up the potential of using outdoor cameras for automated analysis.
Authors: Han Ling
A major challenge faced by current optical flow methods is the difficulty in generalizing them well into the real world, mainly due to the high production cost of datasets, which currently do not have a large real-world optical flow dataset. To address this challenge, we introduce a novel optical flow training framework that can efficiently train optical flow networks on the target data domain without manual annotation. Specifically, we use advanced Nerf technology to reconstruct scenes from photo groups collected by monocular cameras, and calculate the optical flow results between camera pose pairs from the rendered results. On this basis, we screen the generated training data from various aspects such as Nerf's reconstruction quality, visual consistency of optical flow labels, reconstruction depth consistency, etc. The filtered training data can be directly used for network supervision. Experimentally, the generalization ability of our scheme on KITTI surpasses existing self-supervised optical flow and monocular scene flow algorithms. Moreover, it can always surpass most supervised methods in real-world zero-point generalization evaluation.
Authors: Chau Pham, Piotr Teterwak, Soren Nelson, Bryan A. Plummer
Most deep neural networks are trained under fixed network architectures and require retraining when the architecture changes. If expanding the network's size is needed, it is necessary to retrain from scratch, which is expensive. To avoid this, one can grow from a small network by adding random weights over time to gradually achieve the target network size. However, this naive approach falls short in practice as it brings too much noise to the growing process. Prior work tackled this issue by leveraging the already learned weights and training data for generating new weights through conducting a computationally expensive analysis step. In this paper, we introduce MixtureGrowth, a new approach to growing networks that circumvents the initialization overhead in prior work. Before growing, each layer in our model is generated with a linear combination of parameter templates. Newly grown layer weights are generated by using a new linear combination of existing templates for a layer. On one hand, these templates are already trained for the task, providing a strong initialization. On the other, the new coefficients provide flexibility for the added layer weights to learn something new. We show that our approach boosts top-1 accuracy over the state-of-the-art by 2-2.5% on CIFAR-100 and ImageNet datasets, while achieving comparable performance with fewer FLOPs to a larger network trained from scratch. Code is available at https://github.com/chaudatascience/mixturegrowth.
Authors: Qinghao Ye, Haiyang Xu, Jiabo Ye, Ming Yan, Haowei Liu, Qi Qian, Ji Zhang, Fei Huang, Jingren Zhou
Multi-modal Large Language Models (MLLMs) have demonstrated impressive instruction abilities across various open-ended tasks. However, previous methods primarily focus on enhancing multi-modal capabilities. In this work, we introduce a versatile multi-modal large language model, mPLUG-Owl2, which effectively leverages modality collaboration to improve performance in both text and multi-modal tasks. mPLUG-Owl2 utilizes a modularized network design, with the language decoder acting as a universal interface for managing different modalities. Specifically, mPLUG-Owl2 incorporates shared functional modules to facilitate modality collaboration and introduces a modality-adaptive module that preserves modality-specific features. Extensive experiments reveal that mPLUG-Owl2 is capable of generalizing both text tasks and multi-modal tasks and achieving state-of-the-art performances with a single generic model. Notably, mPLUG-Owl2 is the first MLLM model that demonstrates the modality collaboration phenomenon in both pure-text and multi-modal scenarios, setting a pioneering path in the development of future multi-modal foundation models.
Authors: Motonari Kambara, Komei Sugiura
This paper aims to develop a framework that enables a robot to execute tasks based on visual information, in response to natural language instructions for Fetch-and-Carry with Object Grounding (FCOG) tasks. Although there have been many frameworks, they usually rely on manually given instruction sentences. Therefore, evaluations have only been conducted with fixed tasks. Furthermore, many multimodal language understanding models for the benchmarks only consider discrete actions. To address the limitations, we propose a framework for the full automation of the generation, execution, and evaluation of FCOG tasks. In addition, we introduce an approach to solving the FCOG tasks by dividing them into four distinct subtasks.
Authors: Lorenzo Agnolucci, Leonardo Galteri, Marco Bertini, Alberto Del Bimbo
In this paper, we present a system to restore analog videos of historical archives. These videos often contain severe visual degradation due to the deterioration of their tape supports that require costly and slow manual interventions to recover the original content. The proposed system uses a multi-frame approach and is able to deal with severe tape mistracking, which results in completely scrambled frames. Tests on real-world videos from a major historical video archive show the effectiveness of our demo system. The code and the pre-trained model are publicly available at https://github.com/miccunifi/analog-video-restoration.
Authors: Muntabir Hasan Choudhury, Lamia Salsabil, William A. Ingram, Edward A. Fox, Jian Wu
Electronic theses and dissertations (ETDs) have been proposed, advocated, and generated for more than 25 years. Although ETDs are hosted by commercial or institutional digital library repositories, they are still an understudied type of scholarly big data, partially because they are usually longer than conference proceedings and journals. Segmenting ETDs will allow researchers to study sectional content. Readers can navigate to particular pages of interest, discover, and explore the content buried in these long documents. Most existing frameworks on document page classification are designed for classifying general documents and perform poorly on ETDs. In this paper, we propose ETDPC. Its backbone is a two-stream multimodal model with a cross-attention network to classify ETD pages into 13 categories. To overcome the challenge of imbalanced labeled samples, we augmented data for minority categories and employed a hierarchical classifier. ETDPC outperforms the state-of-the-art models in all categories, achieving an F1 of 0.84 -- 0.96 for 9 out of 13 categories. We also demonstrated its data efficiency. The code and data can be found on GitHub (https://github.com/lamps-lab/ETDMiner/tree/master/etd_segmentation).
Authors: Lorenzo Agnolucci, Leonardo Galteri, Marco Bertini, Alberto Del Bimbo
In the latest years, videoconferencing has taken a fundamental role in interpersonal relations, both for personal and business purposes. Lossy video compression algorithms are the enabling technology for videoconferencing, as they reduce the bandwidth required for real-time video streaming. However, lossy video compression decreases the perceived visual quality. Thus, many techniques for reducing compression artifacts and improving video visual quality have been proposed in recent years. In this work, we propose a novel GAN-based method for compression artifacts reduction in videoconferencing. Given that, in this context, the speaker is typically in front of the camera and remains the same for the entire duration of the transmission, we can maintain a set of reference keyframes of the person from the higher-quality I-frames that are transmitted within the video stream and exploit them to guide the visual quality improvement; a novel aspect of this approach is the update policy that maintains and updates a compact and effective set of reference keyframes. First, we extract multi-scale features from the compressed and reference frames. Then, our architecture combines these features in a progressive manner according to facial landmarks. This allows the restoration of the high-frequency details lost after the video compression. Experiments show that the proposed approach improves visual quality and generates photo-realistic results even with high compression rates. Code and pre-trained networks are publicly available at https://github.com/LorenzoAgnolucci/Keyframes-GAN.
Authors: Tony Lee, Michihiro Yasunaga, Chenlin Meng, Yifan Mai, Joon Sung Park, Agrim Gupta, Yunzhi Zhang, Deepak Narayanan, Hannah Benita Teufel, Marco Bellagente, Minguk Kang, Taesung Park, Jure Leskovec, Jun-Yan Zhu, Li Fei-Fei, Jiajun Wu, Stefano Ermon, Percy Liang
The stunning qualitative improvement of recent text-to-image models has led to their widespread attention and adoption. However, we lack a comprehensive quantitative understanding of their capabilities and risks. To fill this gap, we introduce a new benchmark, Holistic Evaluation of Text-to-Image Models (HEIM). Whereas previous evaluations focus mostly on text-image alignment and image quality, we identify 12 aspects, including text-image alignment, image quality, aesthetics, originality, reasoning, knowledge, bias, toxicity, fairness, robustness, multilinguality, and efficiency. We curate 62 scenarios encompassing these aspects and evaluate 26 state-of-the-art text-to-image models on this benchmark. Our results reveal that no single model excels in all aspects, with different models demonstrating different strengths. We release the generated images and human evaluation results for full transparency at https://crfm.stanford.edu/heim/v1.1.0 and the code at https://github.com/stanford-crfm/helm, which is integrated with the HELM codebase.
Authors: Amritpal Singh, Mustafa Burak Gurbuz, Shiva Souhith Gantha, Prahlad Jasti
Healthcare clinics regularly encounter dynamic data that changes due to variations in patient populations, treatment policies, medical devices, and emerging disease patterns. Deep learning models can suffer from catastrophic forgetting when fine-tuned in such scenarios, causing poor performance on previously learned tasks. Continual learning allows learning on new tasks without performance drop on previous tasks. In this work, we investigate the performance of continual learning models on four different medical imaging scenarios involving ten classification datasets from diverse modalities, clinical specialties, and hospitals. We implement various continual learning approaches and evaluate their performance in these scenarios. Our results demonstrate that a single model can sequentially learn new tasks from different specialties and achieve comparable performance to naive methods. These findings indicate the feasibility of recycling or sharing models across the same or different medical specialties, offering another step towards the development of general-purpose medical imaging AI that can be shared across institutions.
Authors: Xingzhe He, Zhiwen Cao, Nicholas Kolkin, Lantao Yu, Helge Rhodin, Ratheesh Kalarot
Large text-to-image models have revolutionized the ability to generate imagery using natural language. However, particularly unique or personal visual concepts, such as your pet, an object in your house, etc., will not be captured by the original model. This has led to interest in how to inject new visual concepts, bound to a new text token, using as few as 4-6 examples. Despite significant progress, this task remains a formidable challenge, particularly in preserving the subject's identity. While most researchers attempt to to address this issue by modifying model architectures, our approach takes a data-centric perspective, advocating the modification of data rather than the model itself. We introduce a novel regularization dataset generation strategy on both the text and image level; demonstrating the importance of a rich and structured regularization dataset (automatically generated) to prevent losing text coherence and better identity preservation. The better quality is enabled by allowing up to 5x more fine-tuning iterations without overfitting and degeneration. The generated renditions of the desired subject preserve even fine details such as text and logos; all while maintaining the ability to generate diverse samples that follow the input text prompt. Since our method focuses on data augmentation, rather than adjusting the model architecture, it is complementary and can be combined with prior work. We show on established benchmarks that our data-centric approach forms the new state of the art in terms of image quality, with the best trade-off between identity preservation, diversity, and text alignment.
Authors: Seungwook Kim, Juhong Min, Minsu Cho
Recent studies show that leveraging the match-wise relationships within the 4D correlation map yields significant improvements in establishing semantic correspondences - but at the cost of increased computation and latency. In this work, we focus on the aspect that the performance improvements of recent methods can also largely be attributed to the usage of multi-scale correlation maps, which hold various information ranging from low-level geometric cues to high-level semantic contexts. To this end, we propose HCCNet, an efficient yet effective semantic matching method which exploits the full potential of multi-scale correlation maps, while eschewing the reliance on expensive match-wise relationship mining on the 4D correlation map. Specifically, HCCNet performs feature slicing on the bottleneck features to yield a richer set of intermediate features, which are used to construct a hypercolumn correlation. HCCNet can consequently establish semantic correspondences in an effective manner by reducing the volume of conventional high-dimensional convolution or self-attention operations to efficient point-wise convolutions. HCCNet demonstrates state-of-the-art or competitive performances on the standard benchmarks of semantic matching, while incurring a notably lower latency and computation overhead compared to the existing SoTA methods.
Authors: Mahdi Ghafourian, Julian Fierrez, Ruben Vera-Rodriguez, Ruben Tolosana, Aythami Morales
Federated Learning (FL) is a machine learning paradigm to conduct collaborative learning among clients on a joint model. The primary goal is to share clients' local training parameters with an integrating server while preserving their privacy. This method permits to exploit the potential of massive mobile users' data for the benefit of machine learning models' performance while keeping sensitive data on local devices. On the downside, FL raises security and privacy concerns that have just started to be studied. To address some of the key threats in FL, researchers have proposed to use secure aggregation methods (e.g. homomorphic encryption, secure multiparty computation, etc.). These solutions improve some security and privacy metrics, but at the same time bring about other serious threats such as poisoning attacks, backdoor attacks, and free running attacks. This paper proposes a new defense method against poisoning attacks in FL called SaFL (Sybil-aware Federated Learning) that minimizes the effect of sybils with a novel time-variant aggregation scheme.
Authors: Gopikrishna Pavuluri, Gayathri Annem
In this research we propose a deep learning approach for detecting anomalies in videos using convolutional autoencoder and decoder neural networks on the UCSD dataset.Our method utilizes a convolutional autoencoder to learn the spatiotemporal patterns of normal videos and then compares each frame of a test video to this learned representation. We evaluated our approach on the UCSD dataset and achieved an overall accuracy of 99.35% on the Ped1 dataset and 99.77% on the Ped2 dataset, demonstrating the effectiveness of our method for detecting anomalies in surveillance videos. The results show that our method outperforms other state-of-the-art methods, and it can be used in real-world applications for video anomaly detection.
Authors: Emmanuel Hartman, Emery Pierson, Martin Bauer, Mohamed Daoudi, Nicolas Charon
This paper introduces a new mathematical and numerical framework for surface analysis derived from the general setting of elastic Riemannian metrics on shape spaces. Traditionally, those metrics are defined over the infinite dimensional manifold of immersed surfaces and satisfy specific invariance properties enabling the comparison of surfaces modulo shape preserving transformations such as reparametrizations. The specificity of the approach we develop is to restrict the space of allowable transformations to predefined finite dimensional bases of deformation fields. These are estimated in a data-driven way so as to emulate specific types of surface transformations observed in a training set. The use of such bases allows to simplify the representation of the corresponding shape space to a finite dimensional latent space. However, in sharp contrast with methods involving e.g. mesh autoencoders, the latent space is here equipped with a non-Euclidean Riemannian metric precisely inherited from the family of aforementioned elastic metrics. We demonstrate how this basis restricted model can be then effectively implemented to perform a variety of tasks on surface meshes which, importantly, does not assume these to be pre-registered (i.e. with given point correspondences) or to even have a consistent mesh structure. We specifically validate our approach on human body shape and pose data as well as human face scans, and show how it generally outperforms state-of-the-art methods on problems such as shape registration, interpolation, motion transfer or random pose generation.
Authors: Chenfeng Xu, Huan Ling, Sanja Fidler, Or Litany
We present 3DiffTection, a state-of-the-art method for 3D object detection from single images, leveraging features from a 3D-aware diffusion model. Annotating large-scale image data for 3D detection is resource-intensive and time-consuming. Recently, pretrained large image diffusion models have become prominent as effective feature extractors for 2D perception tasks. However, these features are initially trained on paired text and image data, which are not optimized for 3D tasks, and often exhibit a domain gap when applied to the target data. Our approach bridges these gaps through two specialized tuning strategies: geometric and semantic. For geometric tuning, we fine-tune a diffusion model to perform novel view synthesis conditioned on a single image, by introducing a novel epipolar warp operator. This task meets two essential criteria: the necessity for 3D awareness and reliance solely on posed image data, which are readily available (e.g., from videos) and does not require manual annotation. For semantic refinement, we further train the model on target data with detection supervision. Both tuning phases employ ControlNet to preserve the integrity of the original feature capabilities. In the final step, we harness these enhanced capabilities to conduct a test-time prediction ensemble across multiple virtual viewpoints. Through our methodology, we obtain 3D-aware features that are tailored for 3D detection and excel in identifying cross-view point correspondences. Consequently, our model emerges as a powerful 3D detector, substantially surpassing previous benchmarks, e.g., Cube-RCNN, a precedent in single-view 3D detection by 9.43\% in AP3D on the Omni3D-ARkitscene dataset. Furthermore, 3DiffTection showcases robust data efficiency and generalization to cross-domain data.
Authors: Yicong Hong, Kai Zhang, Jiuxiang Gu, Sai Bi, Yang Zhou, Difan Liu, Feng Liu, Kalyan Sunkavalli, Trung Bui, Hao Tan
We propose the first Large Reconstruction Model (LRM) that predicts the 3D model of an object from a single input image within just 5 seconds. In contrast to many previous methods that are trained on small-scale datasets such as ShapeNet in a category-specific fashion, LRM adopts a highly scalable transformer-based architecture with 500 million learnable parameters to directly predict a neural radiance field (NeRF) from the input image. We train our model in an end-to-end manner on massive multi-view data containing around 1 million objects, including both synthetic renderings from Objaverse and real captures from MVImgNet. This combination of a high-capacity model and large-scale training data empowers our model to be highly generalizable and produce high-quality 3D reconstructions from various testing inputs including real-world in-the-wild captures and images from generative models. Video demos and interactable 3D meshes can be found on this website: https://yiconghong.me/LRM/.
Authors: Thanos Delatolas, Vicky Kalogeiton, Dim P. Papadopoulos
Video Object Segmentation (VOS) is crucial for several applications, from video editing to video data generation. Training a VOS model requires an abundance of manually labeled training videos. The de-facto traditional way of annotating objects requires humans to draw detailed segmentation masks on the target objects at each video frame. This annotation process, however, is tedious and time-consuming. To reduce this annotation cost, in this paper, we propose EVA-VOS, a human-in-the-loop annotation framework for video object segmentation. Unlike the traditional approach, we introduce an agent that predicts iteratively both which frame ("What") to annotate and which annotation type ("How") to use. Then, the annotator annotates only the selected frame that is used to update a VOS module, leading to significant gains in annotation time. We conduct experiments on the MOSE and the DAVIS datasets and we show that: (a) EVA-VOS leads to masks with accuracy close to the human agreement 3.5x faster than the standard way of annotating videos; (b) our frame selection achieves state-of-the-art performance; (c) EVA-VOS yields significant performance gains in terms of annotation time compared to all other methods and baselines.
Authors: Dawit Mureja Argaw, Junsik Kim, In So Kweon
Existing video compression (VC) methods primarily aim to reduce the spatial and temporal redundancies between consecutive frames in a video while preserving its quality. In this regard, previous works have achieved remarkable results on videos acquired under specific settings such as instant (known) exposure time and shutter speed which often result in sharp videos. However, when these methods are evaluated on videos captured under different temporal priors, which lead to degradations like motion blur and low frame rate, they fail to maintain the quality of the contents. In this work, we tackle the VC problem in a general scenario where a given video can be blurry due to predefined camera settings or dynamics in the scene. By exploiting the natural trade-off between visual enhancement and data compression, we formulate VC as a min-max optimization problem and propose an effective framework and training strategy to tackle the problem. Extensive experimental results on several benchmark datasets confirm the effectiveness of our method compared to several state-of-the-art VC approaches.
Authors: Junyan Lin, Feng Gao, Xiaocheng Shi, Junyu Dong, Qian Du
Masked image modeling (MIM) is a highly popular and effective self-supervised learning method for image understanding. Existing MIM-based methods mostly focus on spatial feature modeling, neglecting spectral feature modeling. Meanwhile, existing MIM-based methods use Transformer for feature extraction, some local or high-frequency information may get lost. To this end, we propose a spatial-spectral masked auto-encoder (SS-MAE) for HSI and LiDAR/SAR data joint classification. Specifically, SS-MAE consists of a spatial-wise branch and a spectral-wise branch. The spatial-wise branch masks random patches and reconstructs missing pixels, while the spectral-wise branch masks random spectral channels and reconstructs missing channels. Our SS-MAE fully exploits the spatial and spectral representations of the input data. Furthermore, to complement local features in the training stage, we add two lightweight CNNs for feature extraction. Both global and local features are taken into account for feature modeling. To demonstrate the effectiveness of the proposed SS-MAE, we conduct extensive experiments on three publicly available datasets. Extensive experiments on three multi-source datasets verify the superiority of our SS-MAE compared with several state-of-the-art baselines. The source codes are available at \url{https://github.com/summitgao/SS-MAE}.
Authors: Thi-Ngoc-Hanh Le, HuiGuang Huang, Yi-Ru Chen, Tong-Yee Lee
Video holds significance in computer graphics applications. Because of the heterogeneous of digital devices, retargeting videos becomes an essential function to enhance user viewing experience in such applications. In the research of video retargeting, preserving the relevant visual content in videos, avoiding flicking, and processing time are the vital challenges. Extending image retargeting techniques to the video domain is challenging due to the high running time. Prior work of video retargeting mainly utilizes time-consuming preprocessing to analyze frames. Plus, being tolerant of different video content, avoiding important objects from shrinking, and the ability to play with arbitrary ratios are the limitations that need to be resolved in these systems requiring investigation. In this paper, we present an end-to-end RETVI method to retarget videos to arbitrary aspect ratios. We eliminate the computational bottleneck in the conventional approaches by designing RETVI with two modules, content feature analyzer (CFA) and adaptive deforming estimator (ADE). The extensive experiments and evaluations show that our system outperforms previous work in quality and running time. Visit our project website for more results at $\href{this http URL}{this http URL}$.
Authors: Yao Zhu, Yuefeng Chen, Wei Wang, Xiaofeng Mao, Yue Wang, Zhigang Li, Wang lu, Jindong Wang, Xiangyang Ji
Learning generalized representations from limited training samples is crucial for applying deep neural networks in low-resource scenarios. Recently, methods based on Contrastive Language-Image Pre-training (CLIP) have exhibited promising performance in few-shot adaptation tasks. To avoid catastrophic forgetting and overfitting caused by few-shot fine-tuning, existing works usually freeze the parameters of CLIP pre-trained on large-scale datasets, overlooking the possibility that some parameters might not be suitable for downstream tasks. To this end, we revisit CLIP's visual encoder with a specific focus on its distinctive attention pooling layer, which performs a spatial weighted-sum of the dense feature maps. Given that dense feature maps contain meaningful semantic information, and different semantics hold varying importance for diverse downstream tasks (such as prioritizing semantics like ears and eyes in pet classification tasks rather than side mirrors), using the same weighted-sum operation for dense features across different few-shot tasks might not be appropriate. Hence, we propose fine-tuning the parameters of the attention pooling layer during the training process to encourage the model to focus on task-specific semantics. In the inference process, we perform residual blending between the features pooled by the fine-tuned and the original attention pooling layers to incorporate both the few-shot knowledge and the pre-trained CLIP's prior knowledge. We term this method as Semantic-Aware FinE-tuning (SAFE). SAFE is effective in enhancing the conventional few-shot CLIP and is compatible with the existing adapter approach (termed SAFE-A).
Authors: Che-Yung Shen, Jingxi Li, Tianyi Gan, Mona Jarrahi, Aydogan Ozcan
Optical phase conjugation (OPC) is a nonlinear technique used for counteracting wavefront distortions, with various applications ranging from imaging to beam focusing. Here, we present the design of a diffractive wavefront processor to approximate all-optical phase conjugation operation for input fields with phase aberrations. Leveraging deep learning, a set of passive diffractive layers was optimized to all-optically process an arbitrary phase-aberrated coherent field from an input aperture, producing an output field with a phase distribution that is the conjugate of the input wave. We experimentally validated the efficacy of this wavefront processor by 3D fabricating diffractive layers trained using deep learning and performing OPC on phase distortions never seen by the diffractive processor during its training. Employing terahertz radiation, our physical diffractive processor successfully performed the OPC task through a shallow spatially-engineered volume that axially spans tens of wavelengths. In addition to this transmissive OPC configuration, we also created a diffractive phase-conjugate mirror by combining deep learning-optimized diffractive layers with a standard mirror. Given its compact, passive and scalable nature, our diffractive wavefront processor can be used for diverse OPC-related applications, e.g., turbidity suppression and aberration correction, and is also adaptable to different parts of the electromagnetic spectrum, especially those where cost-effective wavefront engineering solutions do not exist.
Authors: Cheng-Yu Chuang, Pooyan Fazli
Video description entails automatically generating coherent natural language sentences that narrate the content of a given video. We introduce CLearViD, a transformer-based model for video description generation that leverages curriculum learning to accomplish this task. In particular, we investigate two curriculum strategies: (1) progressively exposing the model to more challenging samples by gradually applying a Gaussian noise to the video data, and (2) gradually reducing the capacity of the network through dropout during the training process. These methods enable the model to learn more robust and generalizable features. Moreover, CLearViD leverages the Mish activation function, which provides non-linearity and non-monotonicity and helps alleviate the issue of vanishing gradients. Our extensive experiments and ablation studies demonstrate the effectiveness of the proposed model. The results on two datasets, namely ActivityNet Captions and YouCook2, show that CLearViD significantly outperforms existing state-of-the-art models in terms of both accuracy and diversity metrics.
Authors: Puhua Jiang, Mingze Sun, Ruqi Huang
In this paper, we propose a learning-based framework for non-rigid shape registration without correspondence supervision. Traditional shape registration techniques typically rely on correspondences induced by extrinsic proximity, therefore can fail in the presence of large intrinsic deformations. Spectral mapping methods overcome this challenge by embedding shapes into, geometric or learned, high-dimensional spaces, where shapes are easier to align. However, due to the dependency on abstract, non-linear embedding schemes, the latter can be vulnerable with respect to perturbed or alien input. In light of this, our framework takes the best of both worlds. Namely, we deform source mesh towards the target point cloud, guided by correspondences induced by high-dimensional embeddings learned from deep functional maps (DFM). In particular, the correspondences are dynamically updated according to the intermediate registrations and filtered by consistency prior, which prominently robustify the overall pipeline. Moreover, in order to alleviate the requirement of extrinsically aligned input, we train an orientation regressor on a set of aligned synthetic shapes independent of the training shapes for DFM. Empirical results show that, with as few as dozens of training shapes of limited variability, our pipeline achieves state-of-the-art results on several benchmarks of non-rigid point cloud matching, but also delivers high-quality correspondences between unseen challenging shape pairs that undergo both significant extrinsic and intrinsic deformations, in which case neither traditional registration methods nor intrinsic methods work. The code is available at https://github.com/rqhuang88/DFR.
Authors: Hezhen Hu, Xiaoyi Dong, Jianmin Bao, Dongdong Chen, Lu Yuan, Dong Chen, Houqiang Li
Pre-training is playing an increasingly important role in learning generic feature representation for Person Re-identification (ReID). We argue that a high-quality ReID representation should have three properties, namely, multi-level awareness, occlusion robustness, and cross-region invariance. To this end, we propose a simple yet effective pre-training framework, namely PersonMAE, which involves two core designs into masked autoencoders to better serve the task of Person Re-ID. 1) PersonMAE generates two regions from the given image with RegionA as the input and \textit{RegionB} as the prediction target. RegionA is corrupted with block-wise masking to mimic common occlusion in ReID and its remaining visible parts are fed into the encoder. 2) Then PersonMAE aims to predict the whole RegionB at both pixel level and semantic feature level. It encourages its pre-trained feature representations with the three properties mentioned above. These properties make PersonMAE compatible with downstream Person ReID tasks, leading to state-of-the-art performance on four downstream ReID tasks, i.e., supervised (holistic and occluded setting), and unsupervised (UDA and USL setting). Notably, on the commonly adopted supervised setting, PersonMAE with ViT-B backbone achieves 79.8% and 69.5% mAP on the MSMT17 and OccDuke datasets, surpassing the previous state-of-the-art by a large margin of +8.0 mAP, and +5.3 mAP, respectively.
Authors: Ao Zhang, Liming Zhao, Chen-Wei Xie, Yun Zheng, Wei Ji, Tat-Seng Chua
The development of large language models (LLMs) has greatly advanced the field of multimodal understanding, leading to the emergence of large multimodal models (LMMs). In order to enhance the level of visual comprehension, recent studies have equipped LMMs with region-level understanding capabilities by representing object bounding box coordinates as a series of text sequences (pixel2seq). In this paper, we introduce a novel paradigm for object location modeling called pixel2emb method, where we ask the LMM to output the location embeddings and then decoded by different decoders. This paradigm allows for different location formats (such as bounding boxes and masks) to be used in multimodal conversations Furthermore, this kind of embedding based location modeling enables the utilization of existing practices in localization tasks, such as detection and segmentation. In scenarios with limited resources, our pixel2emb demonstrates superior performance compared to existing state-of-the-art (SOTA) approaches in both the location input and output tasks under fair comparison. Leveraging the proposed pixel2emb method, we train an LMM named NExT-Chat and demonstrate its capability of handling multiple tasks like visual grounding, region caption, and grounded reasoning.
Authors: Hao Yang, Haiyang Wang, Di Dai, Liwei Wang
Pre-training is crucial in 3D-related fields such as autonomous driving where point cloud annotation is costly and challenging. Many recent studies on point cloud pre-training, however, have overlooked the issue of incompleteness, where only a fraction of the points are captured by LiDAR, leading to ambiguity during the training phase. On the other hand, images offer more comprehensive information and richer semantics that can bolster point cloud encoders in addressing the incompleteness issue inherent in point clouds. Yet, incorporating images into point cloud pre-training presents its own challenges due to occlusions, potentially causing misalignments between points and pixels. In this work, we propose PRED, a novel image-assisted pre-training framework for outdoor point clouds in an occlusion-aware manner. The main ingredient of our framework is a Birds-Eye-View (BEV) feature map conditioned semantic rendering, leveraging the semantics of images for supervision through neural rendering. We further enhance our model's performance by incorporating point-wise masking with a high mask ratio (95%). Extensive experiments demonstrate PRED's superiority over prior point cloud pre-training methods, providing significant improvements on various large-scale datasets for 3D perception tasks. Codes will be available at https://github.com/PRED4pc/PRED.
Authors: Thorsten Hempel, Magnus Jung, Ahmed A. Abdelrahman, Ayoub Al-Hamadi
Eye contact is a crucial non-verbal interaction modality and plays an important role in our everyday social life. While humans are very sensitive to eye contact, the capabilities of machines to capture a person's gaze are still mediocre. We tackle this challenge and present NITEC, a hand-annotated eye contact dataset for ego-vision interaction. NITEC exceeds existing datasets for ego-vision eye contact in size and variety of demographics, social contexts, and lighting conditions, making it a valuable resource for advancing ego-vision-based eye contact research. Our extensive evaluations on NITEC demonstrate strong cross-dataset performance, emphasizing its effectiveness and adaptability in various scenarios, that allows seamless utilization to the fields of computer vision, human-computer interaction, and social robotics. We make our NITEC dataset publicly available to foster reproducibility and further exploration in the field of ego-vision interaction. https://github.com/thohemp/nitec
Authors: Yuehai Chen
Crowd counting models in highly congested areas confront two main challenges: weak localization ability and difficulty in differentiating between foreground and background, leading to inaccurate estimations. The reason is that objects in highly congested areas are normally small and high-level features extracted by convolutional neural networks are less discriminative to represent small objects. To address these problems, we propose a learning discriminative features framework for crowd counting, which is composed of a masked feature prediction module (MPM) and a supervised pixel-level contrastive learning module (CLM). The MPM randomly masks feature vectors in the feature map and then reconstructs them, allowing the model to learn about what is present in the masked regions and improving the model's ability to localize objects in high-density regions. The CLM pulls targets close to each other and pushes them far away from background in the feature space, enabling the model to discriminate foreground objects from background. Additionally, the proposed modules can be beneficial in various computer vision tasks, such as crowd counting and object detection, where dense scenes or cluttered environments pose challenges to accurate localization. The proposed two modules are plug-and-play, incorporating the proposed modules into existing models can potentially boost their performance in these scenarios.
Authors: Miao Kang, Shengqi Wang, Sanping Zhou, Ke Ye, Jingjing Jiang, Nanning Zheng
Motion forecasting plays a crucial role in autonomous driving, with the aim of predicting the future reasonable motions of traffic agents. Most existing methods mainly model the historical interactions between agents and the environment, and predict multi-modal trajectories in a feedforward process, ignoring potential trajectory changes caused by future interactions between agents. In this paper, we propose a novel Future Feedback Interaction Network (FFINet) to aggregate features the current observations and potential future interactions for trajectory prediction. Firstly, we employ different spatial-temporal encoders to embed the decomposed position vectors and the current position of each scene, providing rich features for the subsequent cross-temporal aggregation. Secondly, the relative interaction and cross-temporal aggregation strategies are sequentially adopted to integrate features in the current fusion module, observation interaction module, future feedback module and global fusion module, in which the future feedback module can enable the understanding of pre-action by feeding the influence of preview information to feedforward prediction. Thirdly, the comprehensive interaction features are further fed into final predictor to generate the joint predicted trajectories of multiple agents. Extensive experimental results show that our FFINet achieves the state-of-the-art performance on Argoverse 1 and Argoverse 2 motion forecasting benchmarks.
Authors: Nishant Jain, Suryansh Kumar, Luc Van Gool
We introduce an improved solution to the neural image-based rendering problem in computer vision. Given a set of images taken from a freely moving camera at train time, the proposed approach could synthesize a realistic image of the scene from a novel viewpoint at test time. The key ideas presented in this paper are (i) Recovering accurate camera parameters via a robust pipeline from unposed day-to-day images is equally crucial in neural novel view synthesis problem; (ii) It is rather more practical to model object's content at different resolutions since dramatic camera motion is highly likely in day-to-day unposed images. To incorporate the key ideas, we leverage the fundamentals of scene rigidity, multi-scale neural scene representation, and single-image depth prediction. Concretely, the proposed approach makes the camera parameters as learnable in a neural fields-based modeling framework. By assuming per view depth prediction is given up to scale, we constrain the relative pose between successive frames. From the relative poses, absolute camera pose estimation is modeled via a graph-neural network-based multiple motion averaging within the multi-scale neural-fields network, leading to a single loss function. Optimizing the introduced loss function provides camera intrinsic, extrinsic, and image rendering from unposed images. We demonstrate, with examples, that for a unified framework to accurately model multiscale neural scene representation from day-to-day acquired unposed multi-view images, it is equally essential to have precise camera-pose estimates within the scene representation framework. Without considering robustness measures in the camera pose estimation pipeline, modeling for multi-scale aliasing artifacts can be counterproductive. We present extensive experiments on several benchmark datasets to demonstrate the suitability of our approach.
Authors: Virginia Fernandez, Walter Hugo Lopez Pinaya, Pedro Borges, Mark S. Graham, Tom Vercauteren, M. Jorge Cardoso
Generative modelling and synthetic data can be a surrogate for real medical imaging datasets, whose scarcity and difficulty to share can be a nuisance when delivering accurate deep learning models for healthcare applications. In recent years, there has been an increased interest in using these models for data augmentation and synthetic data sharing, using architectures such as generative adversarial networks (GANs) or diffusion models (DMs). Nonetheless, the application of synthetic data to tasks such as 3D magnetic resonance imaging (MRI) segmentation remains limited due to the lack of labels associated with the generated images. Moreover, many of the proposed generative MRI models lack the ability to generate arbitrary modalities due to the absence of explicit contrast conditioning. These limitations prevent the user from adjusting the contrast and content of the images and obtaining more generalisable data for training task-specific models. In this work, we propose brainSPADE3D, a 3D generative model for brain MRI and associated segmentations, where the user can condition on specific pathological phenotypes and contrasts. The proposed joint imaging-segmentation generative model is shown to generate high-fidelity synthetic images and associated segmentations, with the ability to combine pathologies. We demonstrate how the model can alleviate issues with segmentation model performance when unexpected pathologies are present in the data.
Authors: Dragos Tantaru, Elisabeta Oneata, Dan Oneata
The remarkable generative capabilities of denoising diffusion models have raised new concerns regarding the authenticity of the images we see every day on the Internet. However, the vast majority of existing deepfake detection models are tested against previous generative approaches (e.g. GAN) and usually provide only a "fake" or "real" label per image. We believe a more informative output would be to augment the per-image label with a localization map indicating which regions of the input have been manipulated. To this end, we frame this task as a weakly-supervised localization problem and identify three main categories of methods (based on either explanations, local scores or attention), which we compare on an equal footing by using the Xception network as the common backbone architecture. We provide a careful analysis of all the main factors that parameterize the design space: choice of method, type of supervision, dataset and generator used in the creation of manipulated images; our study is enabled by constructing datasets in which only one of the components is varied. Our results show that weakly-supervised localization is attainable, with the best performing detection method (based on local scores) being less sensitive to the looser supervision than to the mismatch in terms of dataset or generator.
Authors: Akshit Jindal, Vikram Goyal, Saket Anand, Chetan Arora
Machine Learning (ML) models become vulnerable to Model Stealing Attacks (MSA) when they are deployed as a service. In such attacks, the deployed model is queried repeatedly to build a labelled dataset. This dataset allows the attacker to train a thief model that mimics the original model. To maximize query efficiency, the attacker has to select the most informative subset of data points from the pool of available data. Existing attack strategies utilize approaches like Active Learning and Semi-Supervised learning to minimize costs. However, in the black-box setting, these approaches may select sub-optimal samples as they train only one thief model. Depending on the thief model's capacity and the data it was pretrained on, the model might even select noisy samples that harm the learning process. In this work, we explore the usage of an ensemble of deep learning models as our thief model. We call our attack Army of Thieves(AOT) as we train multiple models with varying complexities to leverage the crowd's wisdom. Based on the ensemble's collective decision, uncertain samples are selected for querying, while the most confident samples are directly included in the training data. Our approach is the first one to utilize an ensemble of thief models to perform model extraction. We outperform the base approaches of existing state-of-the-art methods by at least 3% and achieve a 21% higher adversarial sample transferability than previous work for models trained on the CIFAR-10 dataset.
Authors: Xiaoting Yin, Hao Shi, Jiaan Chen, Ze Wang, Yaozu Ye, Huajian Ni, Kailun Yang, Kaiwei Wang
Human pose estimation is a critical component in autonomous driving and parking, enhancing safety by predicting human actions. Traditional frame-based cameras and videos are commonly applied, yet, they become less reliable in scenarios under high dynamic range or heavy motion blur. In contrast, event cameras offer a robust solution for navigating these challenging contexts. Predominant methodologies incorporate event cameras into learning frameworks by accumulating events into event frames. However, such methods tend to marginalize the intrinsic asynchronous and high temporal resolution characteristics of events. This disregard leads to a loss in essential temporal dimension data, crucial for safety-critical tasks associated with dynamic human activities. To address this issue and to unlock the 3D potential of event information, we introduce two 3D event representations: the Rasterized Event Point Cloud (RasEPC) and the Decoupled Event Voxel (DEV). The RasEPC collates events within concise temporal slices at identical positions, preserving 3D attributes with statistical cues and markedly mitigating memory and computational demands. Meanwhile, the DEV representation discretizes events into voxels and projects them across three orthogonal planes, utilizing decoupled event attention to retrieve 3D cues from the 2D planes. Furthermore, we develop and release EV-3DPW, a synthetic event-based dataset crafted to facilitate training and quantitative analysis in outdoor scenes. On the public real-world DHP19 dataset, our event point cloud technique excels in real-time mobile predictions, while the decoupled event voxel method achieves the highest accuracy. Experiments reveal our proposed 3D representation methods' superior generalization capacities against traditional RGB images and event frame techniques. Our code and dataset are available at https://github.com/MasterHow/EventPointPose.
Authors: Suryaka Suresh, Bishshoy Das, Vinayak Abrol, Sumantra Dutta Roy
We study how the topology of feature embedding space changes as it passes through the layers of a well-trained deep neural network (DNN) through Betti numbers. Motivated by existing studies using simplicial complexes on shallow fully connected networks (FCN), we present an extended analysis using Cubical homology instead, with a variety of popular deep architectures and real image datasets. We demonstrate that as depth increases, a topologically complicated dataset is transformed into a simple one, resulting in Betti numbers attaining their lowest possible value. The rate of decay in topological complexity (as a metric) helps quantify the impact of architectural choices on the generalization ability. Interestingly from a representation learning perspective, we highlight several invariances such as topological invariance of (1) an architecture on similar datasets; (2) embedding space of a dataset for architectures of variable depth; (3) embedding space to input resolution/size, and (4) data sub-sampling. In order to further demonstrate the link between expressivity \& the generalization capability of a network, we consider the task of ranking pre-trained models for downstream classification task (transfer learning). Compared to existing approaches, the proposed metric has a better correlation to the actually achievable accuracy via fine-tuning the pre-trained model.
Authors: Dominic De Jonge
We introduce LuminanceL1Loss, a novel loss function designed to enhance the performance of image restoration tasks. We demonstrate its superiority over MSE when applied to the Retinexformer, BUIFD and DnCNN architectures. Our proposed LuminanceL1Loss leverages a unique approach by transforming images into grayscale and subsequently computing the MSE loss for both grayscale and color channels. Experimental results demonstrate that this innovative loss function consistently outperforms traditional methods, showcasing its potential in image denoising and other related tasks in image reconstruction. It demonstrates gains up to 4.7dB. The results presented in this study highlight the efficacy of LuminanceL1Loss for various image restoration tasks.
Authors: Rui She, Qiyu Kang, Sijie Wang, Wee Peng Tay, Yong Liang Guan, Diego Navarro Navarro, Andreas Hartmannsgruber
Matching landmark patches from a real-time image captured by an on-vehicle camera with landmark patches in an image database plays an important role in various computer perception tasks for autonomous driving. Current methods focus on local matching for regions of interest and do not take into account spatial neighborhood relationships among the image patches, which typically correspond to objects in the environment. In this paper, we construct a spatial graph with the graph vertices corresponding to patches and edges capturing the spatial neighborhood information. We propose a joint feature and metric learning model with graph-based learning. We provide a theoretical basis for the graph-based loss by showing that the information distance between the distributions conditioned on matched and unmatched pairs is maximized under our framework. We evaluate our model using several street-scene datasets and demonstrate that our approach achieves state-of-the-art matching results.
Authors: Marta Gomez-Barrero, Javier Galbally, Christian Rathgeb, Christoph Busch
The wide deployment of biometric recognition systems in the last two decades has raised privacy concerns regarding the storage and use of biometric data. As a consequence, the ISO/IEC 24745 international standard on biometric information protection has established two main requirements for protecting biometric templates: irreversibility and unlinkability. Numerous efforts have been directed to the development and analysis of irreversible templates. However, there is still no systematic quantitative manner to analyse the unlinkability of such templates. In this paper we address this shortcoming by proposing a new general framework for the evaluation of biometric templates' unlinkability. To illustrate the potential of the approach, it is applied to assess the unlinkability of four state-of-the-art techniques for biometric template protection: biometric salting, Bloom filters, Homomorphic Encryption and block re-mapping. For the last technique, the proposed framework is compared with other existing metrics to show its advantages.
Authors: Linus Franke, Darius Rückert, Laura Fink, Matthias Innmann, Marc Stamminger
In the last few years, deep neural networks opened the doors for big advances in novel view synthesis. Many of these approaches are based on a (coarse) proxy geometry obtained by structure from motion algorithms. Small deficiencies in this proxy can be fixed by neural rendering, but larger holes or missing parts, as they commonly appear for thin structures or for glossy regions, still lead to distracting artifacts and temporal instability. In this paper, we present a novel neural-rendering-based approach to detect and fix such deficiencies. As a proxy, we use a point cloud, which allows us to easily remove outlier geometry and to fill in missing geometry without complicated topological operations. Keys to our approach are (i) a differentiable, blending point-based renderer that can blend out redundant points, as well as (ii) the concept of Visual Error Tomography (VET), which allows us to lift 2D error maps to identify 3D-regions lacking geometry and to spawn novel points accordingly. Furthermore, (iii) by adding points as nested environment maps, our approach allows us to generate high-quality renderings of the surroundings in the same pipeline. In our results, we show that our approach can improve the quality of a point cloud obtained by structure from motion and thus increase novel view synthesis quality significantly. In contrast to point growing techniques, the approach can also fix large-scale holes and missing thin structures effectively. Rendering quality outperforms state-of-the-art methods and temporal stability is significantly improved, while rendering is possible at real-time frame rates.
Authors: Daniil Kirilenko, Vitaliy Vorobyov, Alexey K. Kovalev, Aleksandr I. Panov
Object-centric architectures usually apply a differentiable module to the entire feature map to decompose it into sets of entity representations called slots. Some of these methods structurally resemble clustering algorithms, where the cluster's center in latent space serves as a slot representation. Slot Attention is an example of such a method, acting as a learnable analog of the soft k-means algorithm. Our work employs a learnable clustering method based on the Gaussian Mixture Model. Unlike other approaches, we represent slots not only as centers of clusters but also incorporate information about the distance between clusters and assigned vectors, leading to more expressive slot representations. Our experiments demonstrate that using this approach instead of Slot Attention improves performance in object-centric scenarios, achieving state-of-the-art results in the set property prediction task.
Authors: Biqi Yang, Weiliang Tang, Xiaojie Gao, Xianzhi Li, Yun-Hui Liu, Chi-Wing Fu, Pheng-Ann Heng
In large-scale storehouses, precise instance masks are crucial for robotic bin picking but are challenging to obtain. Existing instance segmentation methods typically rely on a tedious process of scene collection, mask annotation, and network fine-tuning for every single Stock Keeping Unit (SKU). This paper presents SKU-Patch, a new patch-guided instance segmentation solution, leveraging only a few image patches for each incoming new SKU to predict accurate and robust masks, without tedious manual effort and model re-training. Technical-wise, we design a novel transformer-based network with (i) a patch-image correlation encoder to capture multi-level image features calibrated by patch information and (ii) a patch-aware transformer decoder with parallel task heads to generate instance masks. Extensive experiments on four storehouse benchmarks manifest that SKU-Patch is able to achieve the best performance over the state-of-the-art methods. Also, SKU-Patch yields an average of nearly 100% grasping success rate on more than 50 unseen SKUs in a robot-aided auto-store logistic pipeline, showing its effectiveness and practicality.
Authors: Watkinson Gabriel, Cohen Ethan, Bourriez Nicolas, Bendidi Ihab, Bollot Guillaume, Genovesio Auguste
With the surge in available data from various modalities, there is a growing need to bridge the gap between different data types. In this work, we introduce a novel approach to learn cross-modal representations between image data and molecular representations for drug discovery. We propose EMM and IMM, two innovative loss functions built on top of CLIP that leverage weak supervision and cross sites replicates in High-Content Screening. Evaluating our model against known baseline on cross-modal retrieval, we show that our proposed approach allows to learn better representations and mitigate batch effect. In addition, we also present a preprocessing method for the JUMP-CP dataset that effectively reduce the required space from 85Tb to a mere usable 7Tb size, still retaining all perturbations and most of the information content.
Authors: Cathrin Elich, Lukas Kirchdorfer, Jan M. Köhler, Lukas Schott
While multi-task learning (MTL) has gained significant attention in recent years, its underlying mechanisms remain poorly understood. Recent methods did not yield consistent performance improvements over single task learning (STL) baselines, underscoring the importance of gaining more profound insights about challenges specific to MTL. In our study, we challenge common assumptions in MTL in the context of STL: First, the choice of optimizer has only been mildly investigated in MTL. We show the pivotal role of common STL tools such as the Adam optimizer in MTL. We deduce the effectiveness of Adam to its partial loss-scale invariance. Second, the notion of gradient conflicts has often been phrased as a specific problem in MTL. We delve into the role of gradient conflicts in MTL and compare it to STL. For angular gradient alignment we find no evidence that this is a unique problem in MTL. We emphasize differences in gradient magnitude as the main distinguishing factor. Lastly, we compare the transferability of features learned through MTL and STL on common image corruptions, and find no conclusive evidence that MTL leads to superior transferability. Overall, we find surprising similarities between STL and MTL suggesting to consider methods from both fields in a broader context.
Authors: Jianchao Ci, Xin Wang, David Rapado-Rincón, Akshay K. Burusa, Gert Kootstra
Greenhouse production of fruits and vegetables in developed countries is challenged by labor 12 scarcity and high labor costs. Robots offer a good solution for sustainable and cost-effective 13 production. Acquiring accurate spatial information about relevant plant parts is vital for 14 successful robot operation. Robot perception in greenhouses is challenging due to variations in 15 plant appearance, viewpoints, and illumination. This paper proposes a keypoint-detection-based 16 method using data from an RGB-D camera to estimate the 3D pose of peduncle nodes, which 17 provides essential information to harvest the tomato bunches. 18 19 Specifically, this paper proposes a method that detects four anatomical landmarks in the color 20 image and then integrates 3D point-cloud information to determine the 3D pose. A 21 comprehensive evaluation was conducted in a commercial greenhouse to gain insight into the 22 performance of different parts of the method. The results showed: (1) high accuracy in object 23 detection, achieving an Average Precision (AP) of AP@0.5=0.96; (2) an average Percentage of 24 Detected Joints (PDJ) of the keypoints of PhDJ@0.2=94.31%; and (3) 3D pose estimation 25 accuracy with mean absolute errors (MAE) of 11.38o and 9.93o for the relative upper and lower 26 angles between the peduncle and main stem, respectively. Furthermore, the capability to handle 27 variations in viewpoint was investigated, demonstrating the method was robust to view changes. 28 However, canonical and higher views resulted in slightly higher performance compared to other 29 views. Although tomato was selected as a use case, the proposed method is also applicable to 30 other greenhouse crops like pepper.
Authors: Calvin Metzger
Contrastive Language-Image Pretraining (CLIP) models are able to capture the semantic relationship of images and texts and have enabled a wide range of applications, from image retrieval to classification. These models are trained with datasets extracted from web crawls, which are of large quantity but limited quality. This paper explores whether limited amounts higher quality data in a specific domain improve the general performance of CLIP models. To this purpose, we extract text-image data from scientific papers hosted in the arXiv and PubMed Central repositories. Experiments on small-scale CLIP models (ViT B/32) show that model performance increases on average, but only moderately. This result indicates that using the data sources considered in the paper to train large-scale CLIP models is a worthwile research direction.
Authors: Wentao Zhu, Jason Qin, Yuke Lou, Hang Ye, Xiaoxuan Ma, Hai Ci, Yizhou Wang
Humans exhibit a remarkable capacity for anticipating the actions of others and planning their own actions accordingly. In this study, we strive to replicate this ability by addressing the social motion prediction problem. We introduce a new benchmark, a novel formulation, and a cognition-inspired framework. We present Wusi, a 3D multi-person motion dataset under the context of team sports, which features intense and strategic human interactions and diverse pose distributions. By reformulating the problem from a multi-agent reinforcement learning perspective, we incorporate behavioral cloning and generative adversarial imitation learning to boost learning efficiency and generalization. Furthermore, we take into account the cognitive aspects of the human social action planning process and develop a cognitive hierarchy framework to predict strategic human social interactions. We conduct comprehensive experiments to validate the effectiveness of our proposed dataset and approach. Code and data are available at https://walter0807.github.io/Social-CH/.
Authors: Guinan Su, Yanwu Yang, Zhifeng Li
In recent years, audio-driven 3D facial animation has gained significant attention, particularly in applications such as virtual reality, gaming, and video conferencing. However, accurately modeling the intricate and subtle dynamics of facial expressions remains a challenge. Most existing studies approach the facial animation task as a single regression problem, which often fail to capture the intrinsic inter-modal relationship between speech signals and 3D facial animation and overlook their inherent consistency. Moreover, due to the limited availability of 3D-audio-visual datasets, approaches learning with small-size samples have poor generalizability that decreases the performance. To address these issues, in this study, we propose a cross-modal dual-learning framework, termed DualTalker, aiming at improving data usage efficiency as well as relating cross-modal dependencies. The framework is trained jointly with the primary task (audio-driven facial animation) and its dual task (lip reading) and shares common audio/motion encoder components. Our joint training framework facilitates more efficient data usage by leveraging information from both tasks and explicitly capitalizing on the complementary relationship between facial motion and audio to improve performance. Furthermore, we introduce an auxiliary cross-modal consistency loss to mitigate the potential over-smoothing underlying the cross-modal complementary representations, enhancing the mapping of subtle facial expression dynamics. Through extensive experiments and a perceptual user study conducted on the VOCA and BIWI datasets, we demonstrate that our approach outperforms current state-of-the-art methods both qualitatively and quantitatively. We have made our code and video demonstrations available at https://github.com/sabrina-su/iadf.git.
Authors: Haoming Zhuang, Beibei Li, Jingtong Ma, Patrice Monkam, Shouliang Qi, Wei Qian, Dianning He
Background: Ovarian cancer is among the three most frequent gynecologic cancers globally. High-grade serous ovarian cancer (HGSOC) is the most common and aggressive histological type. Guided treatment for HGSOC typically involves platinum-based combination chemotherapy, necessitating an assessment of whether the patient is platinum-resistant. The purpose of this study is to propose a deep learning-based method to determine whether a patient is platinum-resistant using multimodal positron emission tomography/computed tomography (PET/CT) images. Methods: 289 patients with HGSOC were included in this study. An end-to-end SE-SPP-DenseNet model was built by adding Squeeze-Excitation Block (SE Block) and Spatial Pyramid Pooling Layer (SPPLayer) to Dense Convolutional Network (DenseNet). Multimodal data from PET/CT images of the regions of interest (ROI) were used to predict platinum resistance in patients. Results: Through five-fold cross-validation, SE-SPP-DenseNet achieved a high accuracy rate and an area under the curve (AUC) in predicting platinum resistance in patients, which were 92.6% and 0.93, respectively. The importance of incorporating SE Block and SPPLayer into the deep learning model, and considering multimodal data was substantiated by carrying out ablation studies and experiments with single modality data. Conclusions: The obtained classification results indicate that our proposed deep learning framework performs better in predicting platinum resistance in patients, which can help gynecologists make better treatment decisions. Keywords: PET/CT, CNN, SE Block, SPP Layer, Platinum resistance, Ovarian cancer
Authors: Xuhao Shan, Xinyang Li, Ruiquan Ge, Shibin Wu, Ahmed Elazab, Jichao Zhu, Lingyan Zhang, Gangyong Jia, Qingying Xiao, Xiang Wan, Changmiao Wang
Intracerebral Hemorrhage (ICH) is a severe condition resulting from damaged brain blood vessel ruptures, often leading to complications and fatalities. Timely and accurate prognosis and management are essential due to its high mortality rate. However, conventional methods heavily rely on subjective clinician expertise, which can lead to inaccurate diagnoses and delays in treatment. Artificial intelligence (AI) models have been explored to assist clinicians, but many prior studies focused on model modification without considering domain knowledge. This paper introduces a novel deep learning algorithm, GCS-ICHNet, which integrates multimodal brain CT image data and the Glasgow Coma Scale (GCS) score to improve ICH prognosis. The algorithm utilizes a transformer-based fusion module for assessment. GCS-ICHNet demonstrates high sensitivity 81.03% and specificity 91.59%, outperforming average clinicians and other state-of-the-art methods.
Authors: Dinar Sharafutdinov, Stanislav Kuskov, Saian Protasov, Alexey Voropaev
In recent years, computer vision has transformed fields such as medical imaging, object recognition, and geospatial analytics. One of the fundamental tasks in computer vision is semantic image segmentation, which is vital for precise object delineation. Autonomous driving represents one of the key areas where computer vision algorithms are applied. The task of road surface segmentation is crucial in self-driving systems, but it requires a labor-intensive annotation process in several data domains. The work described in this paper aims to improve the efficiency of image segmentation using a convolutional neural network in a multi-sensor setup. This approach leverages lidar (Light Detection and Ranging) annotations to directly train image segmentation models on RGB images. Lidar supplements the images by emitting laser pulses and measuring reflections to provide depth information. However, lidar's sparse point clouds often create difficulties for accurate object segmentation. Segmentation of point clouds requires time-consuming preliminary data preparation and a large amount of computational resources. The key innovation of our approach is the masked loss, addressing sparse ground-truth masks from point clouds. By calculating loss exclusively where lidar points exist, the model learns road segmentation on images by using lidar points as ground truth. This approach allows for blending of different ground-truth data types during model training. Experimental validation of the approach on benchmark datasets shows comparable performance to a high-quality image segmentation model. Incorporating lidar reduces the load on annotations and enables training of image-segmentation models without loss of segmentation quality. The methodology is tested on diverse datasets, both publicly available and proprietary. The strengths and weaknesses of the proposed method are also discussed in the paper.
Authors: Jun-Jee Chao, Selim Engin, Nikhil Chavan-Dafle, Bhoram Lee, Volkan Isler
We study the problem of aligning a video that captures a local portion of an environment to the 2D LiDAR scan of the entire environment. We introduce a method (VioLA) that starts with building a semantic map of the local scene from the image sequence, then extracts points at a fixed height for registering to the LiDAR map. Due to reconstruction errors or partial coverage of the camera scan, the reconstructed semantic map may not contain sufficient information for registration. To address this problem, VioLA makes use of a pre-trained text-to-image inpainting model paired with a depth completion model for filling in the missing scene content in a geometrically consistent fashion to support pose registration. We evaluate VioLA on two real-world RGB-D benchmarks, as well as a self-captured dataset of a large office scene. Notably, our proposed scene completion module improves the pose registration performance by up to 20%.
Authors: Dan Song, Xuanpu Zhang, Juan Zhou, Weizhi Nie, Ruofeng Tong, An-An Liu
Image-based virtual try-on aims to synthesize a naturally dressed person image with a clothing image, which revolutionizes online shopping and inspires related topics within image generation, showing both research significance and commercial potentials. However, there is a great gap between current research progress and commercial applications and an absence of comprehensive overview towards this field to accelerate the development. In this survey, we provide a comprehensive analysis of the state-of-the-art techniques and methodologies in aspects of pipeline architecture, person representation and key modules such as try-on indication, clothing warping and try-on stage. We propose a new semantic criteria with CLIP, and evaluate representative methods with uniformly implemented evaluation metrics on the same dataset. In addition to quantitative and qualitative evaluation of current open-source methods, we also utilize ControlNet to fine-tune a recent large image generation model (PBE) to show future potentials of large-scale models on image-based virtual try-on task. Finally, unresolved issues are revealed and future research directions are prospected to identify key trends and inspire further exploration. The uniformly implemented evaluation metrics, dataset and collected methods will be made public available at https://github.com/little-misfit/Survey-Of-Virtual-Try-On.
Authors: Hubert Baniecki, Maciej Chrabaszcz, Andreas Holzinger, Bastian Pfeifer, Anna Saranti, Przemyslaw Biecek
Evaluating explanations of image classifiers regarding ground truth, e.g. segmentation masks defined by human perception, primarily evaluates the quality of the models under consideration rather than the explanation methods themselves. Driven by this observation, we propose a framework for $\textit{jointly}$ evaluating the robustness of safety-critical systems that $\textit{combine}$ a deep neural network with an explanation method. These are increasingly used in real-world applications like medical image analysis or robotics. We introduce a fine-tuning procedure to (mis)align model$\unicode{x2013}$explanation pipelines with ground truth and use it to quantify the potential discrepancy between worst and best-case scenarios of human alignment. Experiments across various model architectures and post-hoc local interpretation methods provide insights into the robustness of vision transformers and the overall vulnerability of such AI systems to potential adversarial attacks.
Authors: Muhammad Akhtar Munir, Muhammad Haris Khan, M. Saquib Sarfraz, Mohsen Ali
Deep learning based object detectors struggle generalizing to a new target domain bearing significant variations in object and background. Most current methods align domains by using image or instance-level adversarial feature alignment. This often suffers due to unwanted background and lacks class-specific alignment. A straightforward approach to promote class-level alignment is to use high confidence predictions on unlabeled domain as pseudo-labels. These predictions are often noisy since model is poorly calibrated under domain shift. In this paper, we propose to leverage model's predictive uncertainty to strike the right balance between adversarial feature alignment and class-level alignment. We develop a technique to quantify predictive uncertainty on class assignments and bounding-box predictions. Model predictions with low uncertainty are used to generate pseudo-labels for self-training, whereas the ones with higher uncertainty are used to generate tiles for adversarial feature alignment. This synergy between tiling around uncertain object regions and generating pseudo-labels from highly certain object regions allows capturing both image and instance-level context during the model adaptation. We report thorough ablation study to reveal the impact of different components in our approach. Results on five diverse and challenging adaptation scenarios show that our approach outperforms existing state-of-the-art methods with noticeable margins.
Authors: Matt Gorbett, Hossein Shirazi, Indrakshi Ray
Learning from the collective knowledge of data dispersed across private sources can provide neural networks with enhanced generalization capabilities. Federated learning, a method for collaboratively training a machine learning model across remote clients, achieves this by combining client models via the orchestration of a central server. However, current approaches face two critical limitations: i) they struggle to converge when client domains are sufficiently different, and ii) current aggregation techniques produce an identical global model for each client. In this work, we address these issues by reformulating the typical federated learning setup: rather than learning a single global model, we learn N models each optimized for a common objective. To achieve this, we apply a weighted distance minimization to model parameters shared in a peer-to-peer topology. The resulting framework, Iterative Parameter Alignment, applies naturally to the cross-silo setting, and has the following properties: (i) a unique solution for each participant, with the option to globally converge each model in the federation, and (ii) an optional early-stopping mechanism to elicit fairness among peers in collaborative learning settings. These characteristics jointly provide a flexible new framework for iteratively learning from peer models trained on disparate datasets. We find that the technique achieves competitive results on a variety of data partitions compared to state-of-the-art approaches. Further, we show that the method is robust to divergent domains (i.e. disjoint classes across peers) where existing approaches struggle.
Authors: Rohit Venkata Sai Dulam, Chandra Kambhamettu
Developing a new Salient Object Detection (SOD) model involves selecting an ImageNet pre-trained backbone and creating novel feature refinement modules to use backbone features. However, adding new components to a pre-trained backbone needs retraining the whole network on the ImageNet dataset, which requires significant time. Hence, we explore developing a neural network from scratch directly trained on SOD without ImageNet pre-training. Such a formulation offers full autonomy to design task-specific components. To that end, we propose SODAWideNet, an encoder-decoder-style network for Salient Object Detection. We deviate from the commonly practiced paradigm of narrow and deep convolutional models to a wide and shallow architecture, resulting in a parameter-efficient deep neural network. To achieve a shallower network, we increase the receptive field from the beginning of the network using a combination of dilated convolutions and self-attention. Therefore, we propose Multi Receptive Field Feature Aggregation Module (MRFFAM) that efficiently obtains discriminative features from farther regions at higher resolutions using dilated convolutions. Next, we propose Multi-Scale Attention (MSA), which creates a feature pyramid and efficiently computes attention across multiple resolutions to extract global features from larger feature maps. Finally, we propose two variants, SODAWideNet-S (3.03M) and SODAWideNet (9.03M), that achieve competitive performance against state-of-the-art models on five datasets.
Authors: Helena Montenegro, Jaime S. Cardoso
Case-based explanations are an intuitive method to gain insight into the decision-making process of deep learning models in clinical contexts. However, medical images cannot be shared as explanations due to privacy concerns. To address this problem, we propose a novel method for disentangling identity and medical characteristics of images and apply it to anonymize medical images. The disentanglement mechanism replaces some feature vectors in an image while ensuring that the remaining features are preserved, obtaining independent feature vectors that encode the images' identity and medical characteristics. We also propose a model to manufacture synthetic privacy-preserving identities to replace the original image's identity and achieve anonymization. The models are applied to medical and biometric datasets, demonstrating their capacity to generate realistic-looking anonymized images that preserve their original medical content. Additionally, the experiments show the network's inherent capacity to generate counterfactual images through the replacement of medical features.
Authors: Zacharias Anastasakis, Dimitrios Mallis, Markos Diomataris, George Alexandridis, Stefanos Kollias, Vassilis Pitsikalis
We present a novel self-supervised approach for representation learning, particularly for the task of Visual Relationship Detection (VRD). Motivated by the effectiveness of Masked Image Modeling (MIM), we propose Masked Bounding Box Reconstruction (MBBR), a variation of MIM where a percentage of the entities/objects within a scene are masked and subsequently reconstructed based on the unmasked objects. The core idea is that, through object-level masked modeling, the network learns context-aware representations that capture the interaction of objects within a scene and thus are highly predictive of visual object relationships. We extensively evaluate learned representations, both qualitatively and quantitatively, in a few-shot setting and demonstrate the efficacy of MBBR for learning robust visual representations, particularly tailored for VRD. The proposed method is able to surpass state-of-the-art VRD methods on the Predicate Detection (PredDet) evaluation setting, using only a few annotated samples. We make our code available at https://github.com/deeplab-ai/SelfSupervisedVRD.
Authors: Danielle Ferreira, Rima Arnaout
Foundation models are experiencing a surge in popularity. The Segment Anything model (SAM) asserts an ability to segment a wide spectrum of objects but required supervised training at unprecedented scale. We compared SAM's performance (against clinical ground truth) and resources (labeling time, compute) to a modality-specific, label-free self-supervised learning (SSL) method on 25 measurements for 100 cardiac ultrasounds. SAM performed poorly and required significantly more labeling and computing resources, demonstrating worse efficiency than SSL.
Authors: Quentin Bouniot
In this thesis, we develop theoretical, algorithmic and experimental contributions for Machine Learning with limited labels, and more specifically for the tasks of Image Classification and Object Detection in Computer Vision. In a first contribution, we are interested in bridging the gap between theory and practice for popular Meta-Learning algorithms used in Few-Shot Classification. We make connections to Multi-Task Representation Learning, which benefits from solid theoretical foundations, to verify the best conditions for a more efficient meta-learning. Then, to leverage unlabeled data when training object detectors based on the Transformer architecture, we propose both an unsupervised pretraining and a semi-supervised learning method in two other separate contributions. For pretraining, we improve Contrastive Learning for object detectors by introducing the localization information. Finally, our semi-supervised method is the first tailored to transformer-based detectors.
Authors: Yash Jain, Harkirat Behl, Zsolt Kira, Vibhav Vineet
Construction of a universal detector poses a crucial question: How can we most effectively train a model on a large mixture of datasets? The answer lies in learning dataset-specific features and ensembling their knowledge but do all this in a single model. Previous methods achieve this by having separate detection heads on a common backbone but that results in a significant increase in parameters. In this work, we present Mixture-of-Experts as a solution, highlighting that MoEs are much more than a scalability tool. We propose Dataset-Aware Mixture-of-Experts, DAMEX where we train the experts to become an `expert' of a dataset by learning to route each dataset tokens to its mapped expert. Experiments on Universal Object-Detection Benchmark show that we outperform the existing state-of-the-art by average +10.2 AP score and improve over our non-MoE baseline by average +2.0 AP score. We also observe consistent gains while mixing datasets with (1) limited availability, (2) disparate domains and (3) divergent label sets. Further, we qualitatively show that DAMEX is robust against expert representation collapse.
Authors: Timm Hess, Tinne Tuytelaars, Gido M. van de Ven
Recent years have seen considerable progress in the continual training of deep neural networks, predominantly thanks to approaches that add replay or regularization terms to the loss function to approximate the joint loss over all tasks so far. However, we show that even with a perfect approximation to the joint loss, these approaches still suffer from temporary but substantial forgetting when starting to train on a new task. Motivated by this 'stability gap', we propose that continual learning strategies should focus not only on the optimization objective, but also on the way this objective is optimized. While there is some continual learning work that alters the optimization trajectory (e.g., using gradient projection techniques), this line of research is positioned as alternative to improving the optimization objective, while we argue it should be complementary. To evaluate the merits of our proposition, we plan to combine replay-approximated joint objectives with gradient projection-based optimization routines to test whether the addition of the latter provides benefits in terms of (1) alleviating the stability gap, (2) increasing the learning efficiency and (3) improving the final learning outcome.
Authors: Zhenfang Chen, Rui Sun, Wenjun Liu, Yining Hong, Chuang Gan
Recent works have shown that Large Language Models (LLMs) could empower traditional neuro-symbolic models via programming capabilities to translate language into module descriptions, thus achieving strong visual reasoning results while maintaining the model's transparency and efficiency. However, these models usually exhaustively generate the entire code snippet given each new instance of a task, which is extremely ineffective. We propose generative neuro-symbolic visual reasoning by growing and reusing modules. Specifically, our model consists of three unique stages, module initialization, module generation, and module execution. First, given a vision-language task, we adopt LLMs to examine whether we could reuse and grow over established modules to handle this new task. If not, we initialize a new module needed by the task and specify the inputs and outputs of this new module. After that, the new module is created by querying LLMs to generate corresponding code snippets that match the requirements. In order to get a better sense of the new module's ability, we treat few-shot training examples as test cases to see if our new module could pass these cases. If yes, the new module is added to the module library for future reuse. Finally, we evaluate the performance of our model on the testing set by executing the parsed programs with the newly made visual modules to get the results. We find the proposed model possesses several advantages. First, it performs competitively on standard tasks like visual question answering and referring expression comprehension; Second, the modules learned from one task can be seamlessly transferred to new tasks; Last but not least, it is able to adapt to new visual reasoning tasks by observing a few training examples and reusing modules.
Authors: Tianhong Dai, Wei Li, Xilei Cao, Jianzhuang Liu, Xu Jia, Ales Leonardis, Youliang Yan, Shanxin Yuan
High dynamic range (HDR) imaging from multiple low dynamic range (LDR) images has been suffering from ghosting artifacts caused by scene and objects motion. Existing methods, such as optical flow based and end-to-end deep learning based solutions, are error-prone either in detail restoration or ghosting artifacts removal. Comprehensive empirical evidence shows that ghosting artifacts caused by large foreground motion are mainly low-frequency signals and the details are mainly high-frequency signals. In this work, we propose a novel frequency-guided end-to-end deep neural network (FHDRNet) to conduct HDR fusion in the frequency domain, and Discrete Wavelet Transform (DWT) is used to decompose inputs into different frequency bands. The low-frequency signals are used to avoid specific ghosting artifacts, while the high-frequency signals are used for preserving details. Using a U-Net as the backbone, we propose two novel modules: merging module and frequency-guided upsampling module. The merging module applies the attention mechanism to the low-frequency components to deal with the ghost caused by large foreground motion. The frequency-guided upsampling module reconstructs details from multiple frequency-specific components with rich details. In addition, a new RAW dataset is created for training and evaluating multi-frame HDR imaging algorithms in the RAW domain. Extensive experiments are conducted on public datasets and our RAW dataset, showing that the proposed FHDRNet achieves state-of-the-art performance.
Authors: Xiaokun Feng, Xiaotang Chen, Jian Jia, Kaiqi Huang
Sandplay image, as an important psychoanalysis carrier, is a visual scene constructed by the client selecting and placing sand objects (e.g., sand, river, human figures, animals, vegetation, buildings, etc.). As the projection of the client's inner world, it contains high-level semantic information reflecting the client's subjective psychological states, which is different from the common natural image scene that only contains the objective basic semantics (e.g., object's name, attribute, bounding box, etc.). In this work, we take "split" which is a typical psychological semantics related to many emotional and personality problems as the research goal, and we propose an automatic detection model, which can replace the time-consuming and expensive manual analysis process. To achieve that, we design a distribution map generation method projecting the semantic judgment problem into a visual problem, and a feature dimensionality reduction and extraction algorithm which can provide a good representation of split semantics. Besides, we built a sandplay datasets by collecting one sample from each client and inviting 5 therapists to label each sample, which has a large data cost. Experimental results demonstrated the effectiveness of our proposed method.
Authors: Fanwei Kong, Shawn Shadden
Patient-specific cardiac modeling combines geometries of the heart derived from medical images and biophysical simulations to predict various aspects of cardiac function. However, generating simulation-suitable models of the heart from patient image data often requires complicated procedures and significant human effort. We present a fast and automated deep-learning method to construct simulation-suitable models of the heart from medical images. The approach constructs meshes from 3D patient images by learning to deform a small set of deformation handles on a whole heart template. For both 3D CT and MR data, this method achieves promising accuracy for whole heart reconstruction, consistently outperforming prior methods in constructing simulation-suitable meshes of the heart. When evaluated on time-series CT data, this method produced more anatomically and temporally consistent geometries than prior methods, and was able to produce geometries that better satisfy modeling requirements for cardiac flow simulations. Our source code and pretrained networks are available at https://github.com/fkong7/HeartDeformNets.
Authors: Zhongping Zhang, Yiwen Gu, Bryan A. Plummer, Xin Miao, Jiayi Liu, Huayan Wang
Video-based movie genre classification has garnered considerable attention due to its various applications in recommendation systems. Prior work has typically addressed this task by adapting models from traditional video classification tasks, such as action recognition or event detection. However, these models often neglect language elements (e.g., narrations or conversations) present in videos, which can implicitly convey high-level semantics of movie genres, like storylines or background context. Additionally, existing approaches are primarily designed to encode the entire content of the input video, leading to inefficiencies in predicting movie genres. Movie genre prediction may require only a few shots to accurately determine the genres, rendering a comprehensive understanding of the entire video unnecessary. To address these challenges, we propose a Movie genre Classification method based on Language augmentatIon and shot samPling (Movie-CLIP). Movie-CLIP mainly consists of two parts: a language augmentation module to recognize language elements from the input audio, and a shot sampling module to select representative shots from the entire video. We evaluate our method on MovieNet and Condensed Movies datasets, achieving approximate 6-9% improvement in mean Average Precision (mAP) over the baselines. We also generalize Movie-CLIP to the scene boundary detection task, achieving 1.1% improvement in Average Precision (AP) over the state-of-the-art. We release our implementation at github.com/Zhongping-Zhang/Movie-CLIP.
Authors: Chunyu Xie, Heng Cai, Jincheng Li, Fanjing Kong, Xiaoyu Wu, Jianfei Song, Henrique Morimitsu, Lin Yao, Dexin Wang, Xiangzheng Zhang, Dawei Leng, Baochang Zhang, Xiangyang Ji, Yafeng Deng
Vision-language pre-training (VLP) on large-scale datasets has shown premier performance on various downstream tasks. In contrast to plenty of available benchmarks with English corpus, large-scale pre-training datasets and downstream datasets with Chinese corpus remain largely unexplored. In this work, we build a large-scale high-quality Chinese Cross-Modal Benchmark named CCMB for the research community, which contains the currently largest public pre-training dataset Zero and five human-annotated fine-tuning datasets for downstream tasks. Zero contains 250 million images paired with 750 million text descriptions, plus two of the five fine-tuning datasets are also currently the largest ones for Chinese cross-modal downstream tasks. Along with the CCMB, we also develop a VLP framework named R2D2, applying a pre-Ranking + Ranking strategy to learn powerful vision-language representations and a two-way distillation method (i.e., target-guided Distillation and feature-guided Distillation) to further enhance the learning capability. With the Zero and the R2D2 VLP framework, we achieve state-of-the-art performance on twelve downstream datasets from five broad categories of tasks including image-text retrieval, image-text matching, image caption, text-to-image generation, and zero-shot image classification. The datasets, models, and codes are available at https://github.com/yuxie11/R2D2
Authors: Delong Chen, Zhao Wu, Fan Liu, Zaiquan Yang, Huaxi Huang, Ying Tan, Erjin Zhou
Contrastive Language Image Pretraining (CLIP) has received widespread attention, since its learned representations can be transferred well to various downstream tasks. During the training process of the CLIP model, the InfoNCE objective aligns positive image-text pairs and separates negative ones. We show an underlying representation grouping effect during this process: the InfoNCE objective indirectly groups semantically similar representations together via randomly emerged within-modal anchors. Based on this understanding, in this paper, Prototypical Contrastive Language Image Pretraining (ProtoCLIP) is introduced to enhance such grouping by boosting its efficiency and increasing its robustness against the modality gap. Specifically, ProtoCLIP sets up prototype-level discrimination between image and text spaces, which efficiently transfers higher-level structural knowledge. Further, Prototypical Back Translation (PBT) is proposed to decouple representation grouping from representation alignment, resulting in effective learning of meaningful representations under large modality gap. The PBT also enables us to introduce additional external teachers with richer prior language knowledge. ProtoCLIP is trained with an online episodic training strategy, which makes it can be scaled up to unlimited amounts of data. We train our ProtoCLIP on Conceptual Captions and achieved an +5.81% ImageNet linear probing improvement and an +2.01% ImageNet zero-shot classification improvement. On the larger YFCC-15M dataset, ProtoCLIP matches the performance of CLIP with 33% of training time. Codes are available at https://github.com/megvii-research/protoclip.
Authors: William Chen, Siyi Hu, Rajat Talak, Luca Carlone
Abstract semantic 3D scene understanding is a problem of critical importance in robotics. As robots still lack the common-sense knowledge about household objects and locations of an average human, we investigate the use of pre-trained language models to impart common sense for scene understanding. We introduce and compare a wide range of scene classification paradigms that leverage language only (zero-shot, embedding-based, and structured-language) or vision and language (zero-shot and fine-tuned). We find that the best approaches in both categories yield $\sim 70\%$ room classification accuracy, exceeding the performance of pure-vision and graph classifiers. We also find such methods demonstrate notable generalization and transfer capabilities stemming from their use of language.
Authors: Xing Han, Tongzheng Ren, Tan Minh Nguyen, Khai Nguyen, Joydeep Ghosh, Nhat Ho
Recent advances in Transformer architectures have empowered their empirical success in a variety of tasks across different domains. However, existing works mainly focus on predictive accuracy and computational cost, without considering other practical issues, such as robustness to contaminated samples. Recent work by Nguyen et al., (2022) has shown that the self-attention mechanism, which is the center of the Transformer architecture, can be viewed as a non-parametric estimator based on kernel density estimation (KDE). This motivates us to leverage a set of robust kernel density estimation methods for alleviating the issue of data contamination. Specifically, we introduce a series of self-attention mechanisms that can be incorporated into different Transformer architectures and discuss the special properties of each method. We then perform extensive empirical studies on language modeling and image classification tasks. Our methods demonstrate robust performance in multiple scenarios while maintaining competitive results on clean datasets.
Authors: YuanFu Yang, Min Sun
Achieving photorealistic rendering of real-world scenes poses a significant challenge with diverse applications, including mixed reality and virtual reality. Neural networks, extensively explored in solving differential equations, have previously been introduced as implicit representations for photorealistic rendering. However, achieving realism through traditional computing methods is arduous due to the time-consuming optical ray tracing, as it necessitates extensive numerical integration of color, transparency, and opacity values for each sampling point during the rendering process. In this paper, we introduce Quantum Radiance Fields (QRF), which incorporate quantum circuits, quantum activation functions, and quantum volume rendering to represent scenes implicitly. Our results demonstrate that QRF effectively confronts the computational challenges associated with extensive numerical integration by harnessing the parallelism capabilities of quantum computing. Furthermore, current neural networks struggle with capturing fine signal details and accurately modeling high-frequency information and higher-order derivatives. Quantum computing's higher order of nonlinearity provides a distinct advantage in this context. Consequently, QRF leverages two key strengths of quantum computing: highly non-linear processing and extensive parallelism, making it a potent tool for achieving photorealistic rendering of real-world scenes.
Authors: Udbhav Bamba, Neeraj Anand, Saksham Aggarwal, Dilip K. Prasad, Deepak K. Gupta
Binarization is a powerful compression technique for neural networks, significantly reducing FLOPs, but often results in a significant drop in model performance. To address this issue, partial binarization techniques have been developed, but a systematic approach to mixing binary and full-precision parameters in a single network is still lacking. In this paper, we propose a controlled approach to partial binarization, creating a budgeted binary neural network (B2NN) with our MixBin strategy. This method optimizes the mixing of binary and full-precision components, allowing for explicit selection of the fraction of the network to remain binary. Our experiments show that B2NNs created using MixBin outperform those from random or iterative searches and state-of-the-art layer selection methods by up to 3% on the ImageNet-1K dataset. We also show that B2NNs outperform the structured pruning baseline by approximately 23% at the extreme FLOP budget of 15%, and perform well in object tracking, with up to a 12.4% relative improvement over other baselines. Additionally, we demonstrate that B2NNs developed by MixBin can be transferred across datasets, with some cases showing improved performance over directly applying MixBin on the downstream data.
Authors: Indel Pal Singh, Enjie Ghorbel, Anis Kacem, Arunkumar Rathinam, Djamila Aouada
In this paper, a discriminator-free adversarial-based Unsupervised Domain Adaptation (UDA) for Multi-Label Image Classification (MLIC) referred to as DDA-MLIC is proposed. Recently, some attempts have been made for introducing adversarial-based UDA methods in the context of MLIC. However, these methods which rely on an additional discriminator subnet present one major shortcoming. The learning of domain-invariant features may harm their task-specific discriminative power, since the classification and discrimination tasks are decoupled. Herein, we propose to overcome this issue by introducing a novel adversarial critic that is directly deduced from the task-specific classifier. Specifically, a two-component Gaussian Mixture Model (GMM) is fitted on the source and target predictions in order to distinguish between two clusters. This allows extracting a Gaussian distribution for each component. The resulting Gaussian distributions are then used for formulating an adversarial loss based on a Frechet distance. The proposed method is evaluated on several multi-label image datasets covering three different types of domain shift. The obtained results demonstrate that DDA-MLIC outperforms existing state-of-the-art methods in terms of precision while requiring a lower number of parameters. The code is publicly available at github.com/cvi2snt/DDA-MLIC.
Authors: Marco Fiorucci, Peter Naylor, Makoto Yamada
Unsupervised change detection between airborne LiDAR data points, taken at separate times over the same location, can be difficult due to unmatching spatial support and noise from the acquisition system. Most current approaches to detect changes in point clouds rely heavily on the computation of Digital Elevation Models (DEM) images and supervised methods. Obtaining a DEM leads to LiDAR informational loss due to pixelisation, and supervision requires large amounts of labelled data often unavailable in real-world scenarios. We propose an unsupervised approach based on the computation of the transport of 3D LiDAR points over two temporal supports. The method is based on unbalanced optimal transport and can be generalised to any change detection problem with LiDAR data. We apply our approach to publicly available datasets for monitoring urban sprawling in various noise and resolution configurations that mimic several sensors used in practice. Our method allows for unsupervised multi-class classification and outperforms the previous state-of-the-art unsupervised approaches by a significant margin.
Authors: Christopher Lang, Alexander Braun, Lars Schillingmann, Karsten Haug, Abhinav Valada
Self-supervised feature learning enables perception systems to benefit from the vast raw data recorded by vehicle fleets worldwide. While video-level self-supervised learning approaches have shown strong generalizability on classification tasks, the potential to learn dense representations from sequential data has been relatively unexplored. In this work, we propose TempO, a temporal ordering pretext task for pre-training region-level feature representations for perception tasks. We embed each frame by an unordered set of proposal feature vectors, a representation that is natural for object detection or tracking systems, and formulate the sequential ordering by predicting frame transition probabilities in a transformer-based multi-frame architecture whose complexity scales less than quadratic with respect to the sequence length. Extensive evaluations on the BDD100K, nuImages, and MOT17 datasets show that our TempO pre-training approach outperforms single-frame self-supervised learning methods as well as supervised transfer learning initialization strategies, achieving an improvement of +0.7% in mAP for object detection and +2.0% in the HOTA score for multi-object tracking.
Authors: Jamie Wynn, Daniyar Turmukhambetov
Under good conditions, Neural Radiance Fields (NeRFs) have shown impressive results on novel view synthesis tasks. NeRFs learn a scene's color and density fields by minimizing the photometric discrepancy between training views and differentiable renderings of the scene. Once trained from a sufficient set of views, NeRFs can generate novel views from arbitrary camera positions. However, the scene geometry and color fields are severely under-constrained, which can lead to artifacts, especially when trained with few input views.
To alleviate this problem we learn a prior over scene geometry and color, using a denoising diffusion model (DDM). Our DDM is trained on RGBD patches of the synthetic Hypersim dataset and can be used to predict the gradient of the logarithm of a joint probability distribution of color and depth patches. We show that, these gradients of logarithms of RGBD patch priors serve to regularize geometry and color of a scene. During NeRF training, random RGBD patches are rendered and the estimated gradient of the log-likelihood is backpropagated to the color and density fields. Evaluations on LLFF, the most relevant dataset, show that our learned prior achieves improved quality in the reconstructed geometry and improved generalization to novel views. Evaluations on DTU show improved reconstruction quality among NeRF methods.
Authors: Cagri Gungor, Adriana Kovashka
Despite recent attention and exploration of depth for various tasks, it is still an unexplored modality for weakly-supervised object detection (WSOD). We propose an amplifier method for enhancing the performance of WSOD by integrating depth information. Our approach can be applied to any WSOD method based on multiple-instance learning, without necessitating additional annotations or inducing large computational expenses. Our proposed method employs a monocular depth estimation technique to obtain hallucinated depth information, which is then incorporated into a Siamese WSOD network using contrastive loss and fusion. By analyzing the relationship between language context and depth, we calculate depth priors to identify the bounding box proposals that may contain an object of interest. These depth priors are then utilized to update the list of pseudo ground-truth boxes, or adjust the confidence of per-box predictions. Our proposed method is evaluated on six datasets (COCO, PASCAL VOC, Conceptual Captions, Clipart1k, Watercolor2k, and Comic2k) by implementing it on top of two state-of-the-art WSOD methods, and we demonstrate a substantial enhancement in performance.
Authors: Susung Hong, Donghoon Ahn, Seungryong Kim
Existing score-distilling text-to-3D generation techniques, despite their considerable promise, often encounter the view inconsistency problem. One of the most notable issues is the Janus problem, where the most canonical view of an object (\textit{e.g}., face or head) appears in other views. In this work, we explore existing frameworks for score-distilling text-to-3D generation and identify the main causes of the view inconsistency problem -- the embedded bias of 2D diffusion models. Based on these findings, we propose two approaches to debias the score-distillation frameworks for view-consistent text-to-3D generation. Our first approach, called score debiasing, involves cutting off the score estimated by 2D diffusion models and gradually increasing the truncation value throughout the optimization process. Our second approach, called prompt debiasing, identifies conflicting words between user prompts and view prompts using a language model, and adjusts the discrepancy between view prompts and the viewing direction of an object. Our experimental results show that our methods improve the realism of the generated 3D objects by significantly reducing artifacts and achieve a good trade-off between faithfulness to the 2D diffusion models and 3D consistency with little overhead. Our project page is available at~\url{https://susunghong.github.io/Debiased-Score-Distillation-Sampling/}.
Authors: Matteo Toso, Matteo Taiana, Stuart James, Alessio Del Bue
Efficient visual localization is crucial to many applications, such as large-scale deployment of autonomous agents and augmented reality. Traditional visual localization, while achieving remarkable accuracy, relies on extensive 3D models of the scene or large collections of geolocalized images, which are often inefficient to store and to scale to novel environments. In contrast, humans orient themselves using very abstract 2D maps, using the location of clearly identifiable landmarks. Drawing on this and on the success of recent works that explored localization on 2D abstract maps, we propose Flatlandia, a novel visual localization challenge. With Flatlandia, we investigate whether it is possible to localize a visual query by comparing the layout of its common objects detected against the known spatial layout of objects in the map. We formalize the challenge as two tasks at different levels of accuracy to investigate the problem and its possible limitations; for each, we propose initial baseline models and compare them against state-of-the-art 6DoF and 3DoF methods. Code and dataset are publicly available at github.com/IIT-PAVIS/Flatlandia.
Authors: Zhengyuan Jiang, Jinghuai Zhang, Neil Zhenqiang Gong
A generative AI model can generate extremely realistic-looking content, posing growing challenges to the authenticity of information. To address the challenges, watermark has been leveraged to detect AI-generated content. Specifically, a watermark is embedded into an AI-generated content before it is released. A content is detected as AI-generated if a similar watermark can be decoded from it. In this work, we perform a systematic study on the robustness of such watermark-based AI-generated content detection. We focus on AI-generated images. Our work shows that an attacker can post-process a watermarked image via adding a small, human-imperceptible perturbation to it, such that the post-processed image evades detection while maintaining its visual quality. We show the effectiveness of our attack both theoretically and empirically. Moreover, to evade detection, our adversarial post-processing method adds much smaller perturbations to AI-generated images and thus better maintain their visual quality than existing popular post-processing methods such as JPEG compression, Gaussian blur, and Brightness/Contrast. Our work shows the insufficiency of existing watermark-based detection of AI-generated content, highlighting the urgent needs of new methods. Our code is publicly available: https://github.com/zhengyuan-jiang/WEvade.
Authors: Bin Xiao, Murat Simsek, Burak Kantarci, Ala Abu Alkheir
Table Detection has become a fundamental task for visually rich document understanding with the surging number of electronic documents. However, popular public datasets widely used in related studies have inherent limitations, including noisy and inconsistent samples, limited training samples, and limited data sources. These limitations make these datasets unreliable to evaluate the model performance and cannot reflect the actual capacity of models. Therefore, this study revisits some open datasets with high-quality annotations, identifies and cleans the noise, and aligns the annotation definitions of these datasets to merge a larger dataset, termed Open-Tables. Moreover, to enrich the data sources, we propose a new ICT-TD dataset using the PDF files of Information and Communication Technologies (ICT) commodities, a different domain containing unique samples that hardly appear in open datasets. To ensure the label quality of the dataset, we annotated the dataset manually following the guidance of a domain expert. The proposed dataset is challenging and can be a sample of actual cases in the business context. We built strong baselines using various state-of-the-art object detection models. Our experimental results show that the domain differences among existing open datasets are minor despite having different data sources. Our proposed Open-Tables and ICT-TD can provide a more reliable evaluation for models because of their high quality and consistent annotations. Besides, they are more suitable for cross-domain settings. Our experimental results show that in the cross-domain setting, benchmark models trained with cleaned Open-Tables dataset can achieve 0.6\%-2.6\% higher weighted average F1 than the corresponding ones trained with the noisy version of Open-Tables, demonstrating the reliability of the proposed datasets. The datasets are public available.
Authors: Shehzaad Dhuliawala, Mrinmaya Sachan, Carl Allen
We present a latent variable model for classification that provides a novel probabilistic interpretation of neural network softmax classifiers. We derive a variational objective to train the model, analogous to the evidence lower bound (ELBO) used to train variational auto-encoders, that generalises the cross-entropy loss used to train classification models. Treating inputs to the softmax layer as samples of a latent variable, our abstracted perspective reveals a potential inconsistency between their anticipated distribution, required for accurate label predictions to be output, and the empirical distribution found in practice. We augment the variational objective to mitigate such inconsistency and encourage a chosen latent distribution, instead of the implicit assumption in off-the-shelf softmax classifiers. Overall, we provide new theoretical insight into the inner workings of widely-used softmax classification. Empirical evaluation on image and text classification datasets demonstrates that our proposed approach, variational classification, maintains classification accuracy while the reshaped latent space improves other desirable properties of a classifier, such as calibration, adversarial robustness, robustness to distribution shift and sample efficiency useful in low data settings.
Authors: Valentin Barriere, Martin Claverie, Maja Schneider, Guido Lemoine, Raphaël d'Andrimont
Accurate in-season crop type classification is crucial for the crop production estimation and monitoring of agricultural parcels. However, the complexity of the plant growth patterns and their spatio-temporal variability present significant challenges. While current deep learning-based methods show promise in crop type classification from single- and multi-modal time series, most existing methods rely on a single modality, such as satellite optical remote sensing data or crop rotation patterns. We propose a novel approach to fuse multimodal information into a model for improved accuracy and robustness across multiple years and countries. The approach relies on three modalities used: remote sensing time series from Sentinel-2 and Landsat 8 observations, parcel crop rotation and local crop distribution. To evaluate our approach, we release a new annotated dataset of 7.4 million agricultural parcels in France and Netherlands. We associate each parcel with time-series of surface reflectance (Red and NIR) and biophysical variables (LAI, FAPAR). Additionally, we propose a new approach to automatically aggregate crop types into a hierarchical class structure for meaningful model evaluation and a novel data-augmentation technique for early-season classification. Performance of the multimodal approach was assessed at different aggregation level in the semantic domain spanning from 151 to 8 crop types or groups. It resulted in accuracy ranging from 91\% to 95\% for NL dataset and from 85\% to 89\% for FR dataset. Pre-training on a dataset improves domain adaptation between countries, allowing for cross-domain zero-shot learning, and robustness of the performances in a few-shot setting from France to Netherlands. Our proposed approach outperforms comparable methods by enabling learning methods to use the often overlooked spatio-temporal context of parcels, resulting in increased preci...
Authors: Gabriel Kasmi, Laurent Dubus, Yves-Marie Saint Drenan, Philippe Blanc
Neural networks have shown remarkable performance in computer vision, but their deployment in numerous scientific and technical fields is challenging due to their black-box nature. Scientists and practitioners need to evaluate the reliability of a decision, i.e., to know simultaneously if a model relies on the relevant features and whether these features are robust to image corruptions. Existing attribution methods aim to provide human-understandable explanations by highlighting important regions in the image domain, but fail to fully characterize a decision process's reliability. To bridge this gap, we introduce the Wavelet sCale Attribution Method (WCAM), a generalization of attribution from the pixel domain to the space-scale domain using wavelet transforms. Attribution in the wavelet domain reveals where and on what scales the model focuses, thus enabling us to assess whether a decision is reliable. Our code is accessible here: \url{https://github.com/gabrielkasmi/spectral-attribution}.
Authors: Marco Bellagente, Manuel Brack, Hannah Teufel, Felix Friedrich, Björn Deiseroth, Constantin Eichenberg, Andrew Dai, Robert Baldock, Souradeep Nanda, Koen Oostermeijer, Andres Felipe Cruz-Salinas, Patrick Schramowski, Kristian Kersting, Samuel Weinbach
The recent popularity of text-to-image diffusion models (DM) can largely be attributed to the intuitive interface they provide to users. The intended generation can be expressed in natural language, with the model producing faithful interpretations of text prompts. However, expressing complex or nuanced ideas in text alone can be difficult. To ease image generation, we propose MultiFusion that allows one to express complex and nuanced concepts with arbitrarily interleaved inputs of multiple modalities and languages. MutliFusion leverages pre-trained models and aligns them for integration into a cohesive system, thereby avoiding the need for extensive training from scratch. Our experimental results demonstrate the efficient transfer of capabilities from individual modules to the downstream model. Specifically, the fusion of all independent components allows the image generation module to utilize multilingual, interleaved multimodal inputs despite being trained solely on monomodal data in a single language.
Authors: Jinyung Hong, Keun Hee Park, Theodore P. Pavlic
Many interpretable AI approaches have been proposed to provide plausible explanations for a model's decision-making. However, configuring an explainable model that effectively communicates among computational modules has received less attention. A recently proposed shared global workspace theory showed that networks of distributed modules can benefit from sharing information with a bottlenecked memory because the communication constraints encourage specialization, compositionality, and synchronization among the modules. Inspired by this, we propose Concept-Centric Transformers, a simple yet effective configuration of the shared global workspace for interpretability, consisting of: i) an object-centric-based memory module for extracting semantic concepts from input features, ii) a cross-attention mechanism between the learned concept and input embeddings, and iii) standard classification and explanation losses to allow human analysts to directly assess an explanation for the model's classification reasoning. We test our approach against other existing concept-based methods on classification tasks for various datasets, including CIFAR100, CUB-200-2011, and ImageNet, and we show that our model achieves better classification accuracy than all baselines across all problems but also generates more consistent concept-based explanations of classification output.
Authors: Zhaoheng Zheng, Haidong Zhu, Ram Nevatia
In this paper, we study the problem of Compositional Zero-Shot Learning (CZSL), which is to recognize novel attribute-object combinations with pre-existing concepts. Recent researchers focus on applying large-scale Vision-Language Pre-trained (VLP) models like CLIP with strong generalization ability. However, these methods treat the pre-trained model as a black box and focus on pre- and post-CLIP operations, which do not inherently mine the semantic concept between the layers inside CLIP. We propose to dive deep into the architecture and insert adapters, a parameter-efficient technique proven to be effective among large language models, into each CLIP encoder layer. We further equip adapters with concept awareness so that concept-specific features of "object", "attribute", and "composition" can be extracted. We assess our method on four popular CZSL datasets, MIT-States, C-GQA, UT-Zappos, and VAW-CZSL, which shows state-of-the-art performance compared to existing methods on all of them.
Authors: Zhongping Zhang, Jian Zheng, Jacob Zhiyuan Fang, Bryan A. Plummer
Diffusion models have demonstrated impressive performance in text-guided image generation. Current methods that leverage the knowledge of these models for image editing either fine-tune them using the input image (e.g., Imagic) or incorporate structure information as additional constraints (e.g., ControlNet). However, fine-tuning large-scale diffusion models on a single image can lead to severe overfitting issues and lengthy inference time. Information leakage from pretrained models also make it challenging to preserve image content not related to the text input. Additionally, methods that incorporate structural guidance (e.g., edge maps, semantic maps, keypoints) find retaining attributes like colors and textures difficult. Using the input image as a control could mitigate these issues, but since these models are trained via reconstruction, a model can simply hide information about the original image when encoding it to perfectly reconstruct the image without learning the editing task. To address these challenges, we propose a text-to-image editing model with an Image Information Removal module (IIR) that selectively erases color-related and texture-related information from the original image, allowing us to better preserve the text-irrelevant content and avoid issues arising from information hiding. Our experiments on CUB, Outdoor Scenes, and COCO reports our approach achieves the best editability-fidelity trade-off results. In addition, a user study on COCO shows that our edited images are preferred 35% more often than prior work.
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: Pum Jun Kim, Yoojin Jang, Jisu Kim, Jaejun Yoo
We propose a robust and reliable evaluation metric for generative models by introducing topological and statistical treatments for rigorous support estimation. Existing metrics, such as Inception Score (IS), Frechet Inception Distance (FID), and the variants of Precision and Recall (P&R), heavily rely on supports that are estimated from sample features. However, the reliability of their estimation has not been seriously discussed (and overlooked) even though the quality of the evaluation entirely depends on it. In this paper, we propose Topological Precision and Recall (TopP&R, pronounced 'topper'), which provides a systematic approach to estimating supports, retaining only topologically and statistically important features with a certain level of confidence. This not only makes TopP&R strong for noisy features, but also provides statistical consistency. Our theoretical and experimental results show that TopP&R is robust to outliers and non-independent and identically distributed (Non-IID) perturbations, while accurately capturing the true trend of change in samples. To the best of our knowledge, this is the first evaluation metric focused on the robust estimation of the support and provides its statistical consistency under noise.
Authors: Zijian Zhou, Oluwatosin Alabi, Meng Wei, Tom Vercauteren, Miaojing Shi
In this paper, we propose a novel text promptable surgical instrument segmentation approach to overcome challenges associated with diversity and differentiation of surgical instruments in minimally invasive surgeries. We redefine the task as text promptable, thereby enabling a more nuanced comprehension of surgical instruments and adaptability to new instrument types. Inspired by recent advancements in vision-language models, we leverage pretrained image and text encoders as our model backbone and design a text promptable mask decoder consisting of attention- and convolution-based prompting schemes for surgical instrument segmentation prediction. Our model leverages multiple text prompts for each surgical instrument through a new mixture of prompts mechanism, resulting in enhanced segmentation performance. Additionally, we introduce a hard instrument area reinforcement module to improve image feature comprehension and segmentation precision. Extensive experiments on several surgical instrument segmentation datasets demonstrate our model's superior performance and promising generalization capability. To our knowledge, this is the first implementation of a promptable approach to surgical instrument segmentation, offering significant potential for practical application in the field of robotic-assisted surgery. Code is available at https://github.com/franciszzj/TP-SIS.
Authors: Yan Ju, Shu Hu, Shan Jia, George H. Chen, Siwei Lyu
Despite the development of effective deepfake detectors in recent years, recent studies have demonstrated that biases in the data used to train these detectors can lead to disparities in detection accuracy across different races and genders. This can result in different groups being unfairly targeted or excluded from detection, allowing undetected deepfakes to manipulate public opinion and erode trust in a deepfake detection model. While existing studies have focused on evaluating fairness of deepfake detectors, to the best of our knowledge, no method has been developed to encourage fairness in deepfake detection at the algorithm level. In this work, we make the first attempt to improve deepfake detection fairness by proposing novel loss functions that handle both the setting where demographic information (eg, annotations of race and gender) is available as well as the case where this information is absent. Fundamentally, both approaches can be used to convert many existing deepfake detectors into ones that encourages fairness. Extensive experiments on four deepfake datasets and five deepfake detectors demonstrate the effectiveness and flexibility of our approach in improving deepfake detection fairness. Our code is available at https://github.com/littlejuyan/DF_Fairness.
Authors: Cian Eastwood, Shashank Singh, Andrei Liviu Nicolicioiu, Marin Vlastelica, Julius von Kügelgen, Bernhard Schölkopf
To avoid failures on out-of-distribution data, recent works have sought to extract features that have an invariant or stable relationship with the label across domains, discarding "spurious" or unstable features whose relationship with the label changes across domains. However, unstable features often carry complementary information that could boost performance if used correctly in the test domain. In this work, we show how this can be done without test-domain labels. In particular, we prove that pseudo-labels based on stable features provide sufficient guidance for doing so, provided that stable and unstable features are conditionally independent given the label. Based on this theoretical insight, we propose Stable Feature Boosting (SFB), an algorithm for: (i) learning a predictor that separates stable and conditionally-independent unstable features; and (ii) using the stable-feature predictions to adapt the unstable-feature predictions in the test domain. Theoretically, we prove that SFB can learn an asymptotically-optimal predictor without test-domain labels. Empirically, we demonstrate the effectiveness of SFB on real and synthetic data.
Authors: Will Rowan, Patrik Huber, Nick Pears, Andrew Keeling
Accurate 3D face reconstruction from 2D images is an enabling technology with applications in healthcare, security, and creative industries. However, current state-of-the-art methods either rely on supervised training with very limited 3D data or self-supervised training with 2D image data. To bridge this gap, we present a method to generate a large-scale synthesised dataset of 250K photorealistic images and their corresponding shape parameters and depth maps, which we call SynthFace. Our synthesis method conditions Stable Diffusion on depth maps sampled from the FLAME 3D Morphable Model (3DMM) of the human face, allowing us to generate a diverse set of shape-consistent facial images that is designed to be balanced in race and gender. We further propose ControlFace, a deep neural network, trained on SynthFace, which achieves competitive performance on the NoW benchmark, without requiring 3D supervision or manual 3D asset creation. The complete SynthFace dataset will be made publicly available upon publication.
Authors: Yuehai Chen, Jing Yang, Badong Chen, Hua Gang, Shaoyi Du
Counting objects in crowded scenes remains a challenge to computer vision. The current deep learning based approach often formulate it as a Gaussian density regression problem. Such a brute-force regression, though effective, may not consider the annotation displacement properly which arises from the human annotation process and may lead to different distributions. We conjecture that it would be beneficial to consider the annotation displacement in the dense object counting task. To obtain strong robustness against annotation displacement, generalized Gaussian distribution (GGD) function with a tunable bandwidth and shape parameter is exploited to form the learning target point annotation probability map, PAPM. Specifically, we first present a hand-designed PAPM method (HD-PAPM), in which we design a function based on GGD to tolerate the annotation displacement. For end-to-end training, the hand-designed PAPM may not be optimal for the particular network and dataset. An adaptively learned PAPM method (AL-PAPM) is proposed. To improve the robustness to annotation displacement, we design an effective transport cost function based on GGD. The proposed PAPM is capable of integration with other methods. We also combine PAPM with P2PNet through modifying the matching cost matrix, forming P2P-PAPM. This could also improve the robustness to annotation displacement of P2PNet. Extensive experiments show the superiority of our proposed methods.
Authors: Frederic Z. Zhang, Yuhui Yuan, Dylan Campbell, Zhuoyao Zhong, Stephen Gould
Recently, the DETR framework has emerged as the dominant approach for human--object interaction (HOI) research. In particular, two-stage transformer-based HOI detectors are amongst the most performant and training-efficient approaches. However, these often condition HOI classification on object features that lack fine-grained contextual information, eschewing pose and orientation information in favour of visual cues about object identity and box extremities. This naturally hinders the recognition of complex or ambiguous interactions. In this work, we study these issues through visualisations and carefully designed experiments. Accordingly, we investigate how best to re-introduce image features via cross-attention. With an improved query design, extensive exploration of keys and values, and box pair positional embeddings as spatial guidance, our model with enhanced predicate visual context (PViC) outperforms state-of-the-art methods on the HICO-DET and V-COCO benchmarks, while maintaining low training cost.
Authors: Enna Sachdeva, Nakul Agarwal, Suhas Chundi, Sean Roelofs, Jiachen Li, Mykel Kochenderfer, Chiho Choi, Behzad Dariush
The widespread adoption of commercial autonomous vehicles (AVs) and advanced driver assistance systems (ADAS) may largely depend on their acceptance by society, for which their perceived trustworthiness and interpretability to riders are crucial. In general, this task is challenging because modern autonomous systems software relies heavily on black-box artificial intelligence models. Towards this goal, this paper introduces a novel dataset, Rank2Tell, a multi-modal ego-centric dataset for Ranking the importance level and Telling the reason for the importance. Using various close and open-ended visual question answering, the dataset provides dense annotations of various semantic, spatial, temporal, and relational attributes of various important objects in complex traffic scenarios. The dense annotations and unique attributes of the dataset make it a valuable resource for researchers working on visual scene understanding and related fields. Furthermore, we introduce a joint model for joint importance level ranking and natural language captions generation to benchmark our dataset and demonstrate performance with quantitative evaluations.
Authors: Minsu Kim, Giseop Kim, Kyong Hwan Jin, Sunwook Choi
A recent sensor fusion in a Bird's Eye View (BEV) space has shown its utility in various tasks such as 3D detection, map segmentation, etc. However, the approach struggles with inaccurate camera BEV estimation, and a perception of distant areas due to the sparsity of LiDAR points. In this paper, we propose a broad BEV fusion (BroadBEV) that addresses the problems with a spatial synchronization approach of cross-modality. Our strategy aims to enhance camera BEV estimation for a broad-sighted perception while simultaneously improving the completion of LiDAR's sparsity in the entire BEV space. Toward that end, we devise Point-scattering that scatters LiDAR BEV distribution to camera depth distribution. The method boosts the learning of depth estimation of the camera branch and induces accurate location of dense camera features in BEV space. For an effective BEV fusion between the spatially synchronized features, we suggest ColFusion that applies self-attention weights of LiDAR and camera BEV features to each other. Our extensive experiments demonstrate that BroadBEV provides a broad-sighted BEV perception with remarkable performance gains.
Authors: Gabriel Kasmi, Laurent Dubus, Yves-Marie Saint-Drenan, Philippe Blanc
Photovoltaic (PV) energy is crucial for the decarbonization of energy systems. Due to the lack of centralized data, remote sensing of rooftop PV installations is the best option to monitor the evolution of the rooftop PV installed fleet at a regional scale. However, current techniques lack reliability and are notably sensitive to shifts in the acquisition conditions. To overcome this, we leverage the wavelet scale attribution method (WCAM), which decomposes a model's prediction in the space-scale domain. The WCAM enables us to assess on which scales the representation of a PV model rests and provides insights to derive methods that improve the robustness to acquisition conditions, thus increasing trust in deep learning systems to encourage their use for the safe integration of clean energy in electric systems.
Authors: Kevin Lin, Donald Brown, Sana Syed, Adam Greene
Eosinophilic Esophagitis (EoE) is an allergic condition increasing in prevalence. To diagnose EoE, pathologists must find 15 or more eosinophils within a single high-power field (400X magnification). Determining whether or not a patient has EoE can be an arduous process and any medical imaging approaches used to assist diagnosis must consider both efficiency and precision. We propose an improvement of Adorno et al's approach for quantifying eosinphils using deep image segmentation. Our new approach leverages Monte Carlo Dropout, a common approach in deep learning to reduce overfitting, to provide uncertainty quantification on current deep learning models. The uncertainty can be visualized in an output image to evaluate model performance, provide insight to how deep learning algorithms function, and assist pathologists in identifying eosinophils.
Authors: Jianhong Bai, Yuchen Yang, Huanpeng Chu, Hualiang Wang, Zuozhu Liu, Ruizhe Chen, Xiaoxuan He, Lianrui Mu, Chengfei Cai, Haoji Hu
Quantization has emerged as a promising direction for model compression. Recently, data-free quantization has been widely studied as a promising method to avoid privacy concerns, which synthesizes images as an alternative to real training data. Existing methods use classification loss to ensure the reliability of the synthesized images. Unfortunately, even if these images are well-classified by the pre-trained model, they still suffer from low semantics and homogenization issues. Intuitively, these low-semantic images are sensitive to perturbations, and the pre-trained model tends to have inconsistent output when the generator synthesizes an image with poor semantics. To this end, we propose Robustness-Guided Image Synthesis (RIS), a simple but effective method to enrich the semantics of synthetic images and improve image diversity, further boosting the performance of downstream data-free compression tasks. Concretely, we first introduce perturbations on input and model weight, then define the inconsistency metrics at feature and prediction levels before and after perturbations. On the basis of inconsistency on two levels, we design a robustness optimization objective to enhance the semantics of synthetic images. Moreover, we also make our approach diversity-aware by forcing the generator to synthesize images with small correlations in the label space. With RIS, we achieve state-of-the-art performance for various settings on data-free quantization and can be extended to other data-free compression tasks.
Authors: Ziyi Yin, Muchao Ye, Tianrong Zhang, Tianyu Du, Jinguo Zhu, Han Liu, Jinghui Chen, Ting Wang, Fenglong Ma
Vision-Language (VL) pre-trained models have shown their superiority on many multimodal tasks. However, the adversarial robustness of such models has not been fully explored. Existing approaches mainly focus on exploring the adversarial robustness under the white-box setting, which is unrealistic. In this paper, we aim to investigate a new yet practical task to craft image and text perturbations using pre-trained VL models to attack black-box fine-tuned models on different downstream tasks. Towards this end, we propose VLAttack to generate adversarial samples by fusing perturbations of images and texts from both single-modal and multimodal levels. At the single-modal level, we propose a new block-wise similarity attack (BSA) strategy to learn image perturbations for disrupting universal representations. Besides, we adopt an existing text attack strategy to generate text perturbations independent of the image-modal attack. At the multimodal level, we design a novel iterative cross-search attack (ICSA) method to update adversarial image-text pairs periodically, starting with the outputs from the single-modal level. We conduct extensive experiments to attack three widely-used VL pretrained models for six tasks on eight datasets. Experimental results show that the proposed VLAttack framework achieves the highest attack success rates on all tasks compared with state-of-the-art baselines, which reveals a significant blind spot in the deployment of pre-trained VL models. Codes will be released soon.
Authors: Anindya Sarkar, Nathan Jacobs, Yevgeniy Vorobeychik
Visual active search (VAS) has been proposed as a modeling framework in which visual cues are used to guide exploration, with the goal of identifying regions of interest in a large geospatial area. Its potential applications include identifying hot spots of rare wildlife poaching activity, search-and-rescue scenarios, identifying illegal trafficking of weapons, drugs, or people, and many others. State of the art approaches to VAS include applications of deep reinforcement learning (DRL), which yield end-to-end search policies, and traditional active search, which combines predictions with custom algorithmic approaches. While the DRL framework has been shown to greatly outperform traditional active search in such domains, its end-to-end nature does not make full use of supervised information attained either during training, or during actual search, a significant limitation if search tasks differ significantly from those in the training distribution. We propose an approach that combines the strength of both DRL and conventional active search by decomposing the search policy into a prediction module, which produces a geospatial distribution of regions of interest based on task embedding and search history, and a search module, which takes the predictions and search history as input and outputs the search distribution. We develop a novel meta-learning approach for jointly learning the resulting combined policy that can make effective use of supervised information obtained both at training and decision time. Our extensive experiments demonstrate that the proposed representation and meta-learning frameworks significantly outperform state of the art in visual active search on several problem domains.
Authors: Yunsheng Ma, Juanwu Lu, Can Cui, Sicheng Zhao, Xu Cao, Wenqian Ye, Ziran Wang
Vehicle-to-vehicle (V2V) communications have greatly enhanced the perception capabilities of connected and automated vehicles (CAVs) by enabling information sharing to "see through the occlusions", resulting in significant performance improvements. However, developing and training complex multi-agent perception models from scratch can be expensive and unnecessary when existing single-agent models show remarkable generalization capabilities. In this paper, we propose a new framework termed MACP, which equips a single-agent pre-trained model with cooperation capabilities. We approach this objective by identifying the key challenges of shifting from single-agent to cooperative settings, adapting the model by freezing most of its parameters and adding a few lightweight modules. We demonstrate in our experiments that the proposed framework can effectively utilize cooperative observations and outperform other state-of-the-art approaches in both simulated and real-world cooperative perception benchmarks while requiring substantially fewer tunable parameters with reduced communication costs. Our source code is available at https://github.com/PurdueDigitalTwin/MACP.
Authors: Hariharan Ravishankar, Rohan Patil, Vikram Melapudi, Stephan Anzengruber, Parminder Bhatia, Kass-Hout Taha, Pavan Annangi
In this paper, we present SonoSAM - a promptable foundational model for segmenting objects of interest on ultrasound images, followed by state of the art tracking model to perform segmentations on 2D+t and 3D ultrasound datasets. Fine-tuned exclusively on a rich, diverse set of objects from $\approx200$k ultrasound image-mask pairs, SonoSAM demonstrates state-of-the-art performance on $8$ unseen ultrasound data-sets, outperforming competing methods by a significant margin on all metrics of interest. SonoSAM achieves average dice similarity score of $>90\%$ on almost all test data-sets within 2-6 clicks on an average, making it a valuable tool for annotating ultrasound images. We also extend SonoSAM to 3-D (2-D +t) applications and demonstrate superior performance making it a valuable tool for generating dense annotations from ultrasound cine-loops. Further, to increase practical utility of SonoSAM, we propose a two-step process of fine-tuning followed by knowledge distillation to a smaller footprint model without comprising the performance. We present detailed qualitative and quantitative comparisons of SonoSAM with state-of-the-art methods showcasing efficacy of SonoSAM as one of the first reliable, generic foundational model for ultrasound.
Authors: Soumyendu Sarkar, Ashwin Ramesh Babu, Sajad Mousavi, Zachariah Carmichael, Vineet Gundecha, Sahand Ghorbanpour, Ricardo Luna, Gutierrez Antonio Guillen, Avisek Naug
We present a novel framework for generating adversarial benchmarks to evaluate the robustness of image classification models. Our framework allows users to customize the types of distortions to be optimally applied to images, which helps address the specific distortions relevant to their deployment. The benchmark can generate datasets at various distortion levels to assess the robustness of different image classifiers. Our results show that the adversarial samples generated by our framework with any of the image classification models, like ResNet-50, Inception-V3, and VGG-16, are effective and transferable to other models causing them to fail. These failures happen even when these models are adversarially retrained using state-of-the-art techniques, demonstrating the generalizability of our adversarial samples. We achieve competitive performance in terms of net $L_2$ distortion compared to state-of-the-art benchmark techniques on CIFAR-10 and ImageNet; however, we demonstrate our framework achieves such results with simple distortions like Gaussian noise without introducing unnatural artifacts or color bleeds. This is made possible by a model-based reinforcement learning (RL) agent and a technique that reduces a deep tree search of the image for model sensitivity to perturbations, to a one-level analysis and action. The flexibility of choosing distortions and setting classification probability thresholds for multiple classes makes our framework suitable for algorithmic audits.
Authors: Johannes Jakubik, Sujit Roy, C. E. Phillips, Paolo Fraccaro, Denys Godwin, Bianca Zadrozny, Daniela Szwarcman, Carlos Gomes, Gabby Nyirjesy, Blair Edwards, Daiki Kimura, Naomi Simumba, Linsong Chu, S. Karthik Mukkavilli, Devyani Lambhate, Kamal Das, Ranjini Bangalore, Dario Oliveira, Michal Muszynski, Kumar Ankur, Muthukumaran Ramasubramanian, Iksha Gurung, Sam Khallaghi, Hanxi (Steve)Li, Michael Cecil, Maryam Ahmadi, Fatemeh Kordi, Hamed Alemohammad, Manil Maskey, Raghu Ganti, Kommy Weldemariam, Rahul Ramachandran
Significant progress in the development of highly adaptable and reusable Artificial Intelligence (AI) models is expected to have a significant impact on Earth science and remote sensing. Foundation models are pre-trained on large unlabeled datasets through self-supervision, and then fine-tuned for various downstream tasks with small labeled datasets. This paper introduces a first-of-a-kind framework for the efficient pre-training and fine-tuning of foundational models on extensive geospatial data. We have utilized this framework to create Prithvi, a transformer-based geospatial foundational model pre-trained on more than 1TB of multispectral satellite imagery from the Harmonized Landsat-Sentinel 2 (HLS) dataset. Our study demonstrates the efficacy of our framework in successfully fine-tuning Prithvi to a range of Earth observation tasks that have not been tackled by previous work on foundation models involving multi-temporal cloud gap imputation, flood mapping, wildfire scar segmentation, and multi-temporal crop segmentation. Our experiments show that the pre-trained model accelerates the fine-tuning process compared to leveraging randomly initialized weights. In addition, pre-trained Prithvi compares well against the state-of-the-art, e.g., outperforming a conditional GAN model in multi-temporal cloud imputation by up to 5pp (or 5.7%) in the structural similarity index. Finally, due to the limited availability of labeled data in the field of Earth observation, we gradually reduce the quantity of available labeled data for refining the model to evaluate data efficiency and demonstrate that data can be decreased significantly without affecting the model's accuracy. The pre-trained 100 million parameter model and corresponding fine-tuning workflows have been released publicly as open source contributions to the global Earth sciences community through Hugging Face.
Authors: Hai Wang, Xiaoyu Xiang, Yuchen Fan, Jing-Hao Xue
Personalized text-to-image (T2I) synthesis based on diffusion models has attracted significant attention in recent research. However, existing methods primarily concentrate on customizing subjects or styles, neglecting the exploration of global geometry. In this study, we propose an approach that focuses on the customization of 360-degree panoramas, which inherently possess global geometric properties, using a T2I diffusion model. To achieve this, we curate a paired image-text dataset specifically designed for the task and subsequently employ it to fine-tune a pre-trained T2I diffusion model with LoRA. Nevertheless, the fine-tuned model alone does not ensure the continuity between the leftmost and rightmost sides of the synthesized images, a crucial characteristic of 360-degree panoramas. To address this issue, we propose a method called StitchDiffusion. Specifically, we perform pre-denoising operations twice at each time step of the denoising process on the stitch block consisting of the leftmost and rightmost image regions. Furthermore, a global cropping is adopted to synthesize seamless 360-degree panoramas. Experimental results demonstrate the effectiveness of our customized model combined with the proposed StitchDiffusion in generating high-quality 360-degree panoramic images. Moreover, our customized model exhibits exceptional generalization ability in producing scenes unseen in the fine-tuning dataset. Code is available at https://github.com/littlewhitesea/StitchDiffusion.
Authors: Guanfang Dong, Anup Basu
Denoising algorithms play a crucial role in medical image processing and analysis. However, classical denoising algorithms often ignore explanatory and critical medical features preservation, which may lead to misdiagnosis and legal liabilities. In this work, we propose a new denoising method for medical images that not only efficiently removes various types of noise, but also preserves key medical features throughout the process. To achieve this goal, we utilize a gradient-based eXplainable Artificial Intelligence (XAI) approach to design a feature preserving loss function. Our feature preserving loss function is motivated by the characteristic that gradient-based XAI is sensitive to noise. Through backpropagation, medical image features before and after denoising can be kept consistent. We conducted extensive experiments on three available medical image datasets, including synthesized 13 different types of noise and artifacts. The experimental results demonstrate the superiority of our method in terms of denoising performance, model explainability, and generalization.
Authors: Fanwei Kong, Sascha Stocker, Perry S. Choi, Michael Ma, Daniel B. Ennis, Alison Marsden
Congenital heart disease (CHD) encompasses a spectrum of cardiovascular structural abnormalities, often requiring customized treatment plans for individual patients. Computational modeling and analysis of these unique cardiac anatomies can improve diagnosis and treatment planning and may ultimately lead to improved outcomes. Deep learning (DL) methods have demonstrated the potential to enable efficient treatment planning by automating cardiac segmentation and mesh construction for patients with normal cardiac anatomies. However, CHDs are often rare, making it challenging to acquire sufficiently large patient cohorts for training such DL models. Generative modeling of cardiac anatomies has the potential to fill this gap via the generation of virtual cohorts; however, prior approaches were largely designed for normal anatomies and cannot readily capture the significant topological variations seen in CHD patients. Therefore, we propose a type- and shape-disentangled generative approach suitable to capture the wide spectrum of cardiac anatomies observed in different CHD types and synthesize differently shaped cardiac anatomies that preserve the unique topology for specific CHD types. Our DL approach represents generic whole heart anatomies with CHD type-specific abnormalities implicitly using signed distance fields (SDF) based on CHD type diagnosis, which conveniently captures divergent anatomical variations across different types and represents meaningful intermediate CHD states. To capture the shape-specific variations, we then learn invertible deformations to morph the learned CHD type-specific anatomies and reconstruct patient-specific shapes. Our approach has the potential to augment the image-segmentation pairs for rarer CHD types for cardiac segmentation and generate cohorts of CHD cardiac meshes for computational simulation.
Authors: Shen Zheng, Changjie Lu, Srinivasa G. Narasimhan
Rain generation algorithms have the potential to improve the generalization of deraining methods and scene understanding in rainy conditions. However, in practice, they produce artifacts and distortions and struggle to control the amount of rain generated due to a lack of proper constraints. In this paper, we propose an unpaired image-to-image translation framework for generating realistic rainy images. We first introduce a Triangular Probability Similarity (TPS) constraint to guide the generated images toward clear and rainy images in the discriminator manifold, thereby minimizing artifacts and distortions during rain generation. Unlike conventional contrastive learning approaches, which indiscriminately push negative samples away from the anchors, we propose a Semantic Noise Contrastive Estimation (SeNCE) strategy and reassess the pushing force of negative samples based on the semantic similarity between the clear and the rainy images and the feature similarity between the anchor and the negative samples. Experiments demonstrate realistic rain generation with minimal artifacts and distortions, which benefits image deraining and object detection in rain. Furthermore, the method can be used to generate realistic snowy and night images, underscoring its potential for broader applicability. Code is available at https://github.com/ShenZheng2000/TPSeNCE.
Authors: Yuanxin Liu, Lei Li, Shuhuai Ren, Rundong Gao, Shicheng Li, Sishuo Chen, Xu Sun, Lu Hou
Recently, open-domain text-to-video (T2V) generation models have made remarkable progress. However, the promising results are mainly shown by the qualitative cases of generated videos, while the quantitative evaluation of T2V models still faces two critical problems. Firstly, existing studies lack fine-grained evaluation of T2V models on different categories of text prompts. Although some benchmarks have categorized the prompts, their categorization either only focuses on a single aspect or fails to consider the temporal information in video generation. Secondly, it is unclear whether the automatic evaluation metrics are consistent with human standards. To address these problems, we propose FETV, a benchmark for Fine-grained Evaluation of Text-to-Video generation. FETV is multi-aspect, categorizing the prompts based on three orthogonal aspects: the major content, the attributes to control and the prompt complexity. FETV is also temporal-aware, which introduces several temporal categories tailored for video generation. Based on FETV, we conduct comprehensive manual evaluations of four representative T2V models, revealing their pros and cons on different categories of prompts from different aspects. We also extend FETV as a testbed to evaluate the reliability of automatic T2V metrics. The multi-aspect categorization of FETV enables fine-grained analysis of the metrics' reliability in different scenarios. We find that existing automatic metrics (e.g., CLIPScore and FVD) correlate poorly with human evaluation. To address this problem, we explore several solutions to improve CLIPScore and FVD, and develop two automatic metrics that exhibit significant higher correlation with humans than existing metrics. Benchmark page: https://github.com/llyx97/FETV.
Authors: Li Ping Qian, Yi Zhang, Sikai Lyu, Huijie Zhu, Yuan Wu, Xuemin Sherman Shen, Xiaoniu Yang
With the rapid development of Artificial Intelligent Internet of Things (AIoT), the image data from AIoT devices has been witnessing the explosive increasing. In this paper, a novel deep image semantic communication model is proposed for the efficient image communication in AIoT. Particularly, at the transmitter side, a high-precision image semantic segmentation algorithm is proposed to extract the semantic information of the image to achieve significant compression of the image data. At the receiver side, a semantic image restoration algorithm based on Generative Adversarial Network (GAN) is proposed to convert the semantic image to a real scene image with detailed information. Simulation results demonstrate that the proposed image semantic communication model can improve the image compression ratio and recovery accuracy by 71.93% and 25.07% on average in comparison with WebP and CycleGAN, respectively. More importantly, our demo experiment shows that the proposed model reduces the total delay by 95.26% in the image communication, when comparing with the original image transmission.
Authors: Arti Kumbhar, Amruta Chougule, Priya Lokhande, Saloni Navaghane, Aditi Burud, Saee Nimbalkar
Utilizing Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Generative Adversarial Networks (GANs), our system introduces an innovative approach to defect detection in manufacturing. This technology excels in precisely identifying faults by extracting intricate details from product photographs, utilizing RNNs to detect evolving errors and generating synthetic defect data to bolster the model's robustness and adaptability across various defect scenarios. The project leverages a deep learning framework to automate real-time flaw detection in the manufacturing process. It harnesses extensive datasets of annotated images to discern complex defect patterns. This integrated system seamlessly fits into production workflows, thereby boosting efficiency and elevating product quality. As a result, it reduces waste and operational costs, ultimately enhancing market competitiveness.
Authors: Injae Kim, Minhyuk Choi, Hyunwoo J. Kim
Neural Radiance Field (NeRF) has enabled novel view synthesis with high fidelity given images and camera poses. Subsequent works even succeeded in eliminating the necessity of pose priors by jointly optimizing NeRF and camera pose. However, these works are limited to relatively simple settings such as photometrically consistent and occluder-free image collections or a sequence of images from a video. So they have difficulty handling unconstrained images with varying illumination and transient occluders. In this paper, we propose $\textbf{UP-NeRF}$ ($\textbf{U}$nconstrained $\textbf{P}$ose-prior-free $\textbf{Ne}$ural $\textbf{R}$adiance $\textbf{F}$ields) to optimize NeRF with unconstrained image collections without camera pose prior. We tackle these challenges with surrogate tasks that optimize color-insensitive feature fields and a separate module for transient occluders to block their influence on pose estimation. In addition, we introduce a candidate head to enable more robust pose estimation and transient-aware depth supervision to minimize the effect of incorrect prior. Our experiments verify the superior performance of our method compared to the baselines including BARF and its variants in a challenging internet photo collection, $\textit{Phototourism}$ dataset.
Authors: Ruyu Wang, Sabrina Schmedding, Marco F. Huber
Recent deep generative models (DGMs) such as generative adversarial networks (GANs) and diffusion probabilistic models (DPMs) have shown their impressive ability in generating high-fidelity photorealistic images. Although looking appealing to human eyes, training a model on purely synthetic images for downstream image processing tasks like image classification often results in an undesired performance drop compared to training on real data. Previous works have demonstrated that enhancing a real dataset with synthetic images from DGMs can be beneficial. However, the improvements were subjected to certain circumstances and yet were not comparable to adding the same number of real images. In this work, we propose a new taxonomy to describe factors contributing to this commonly observed phenomenon and investigate it on the popular CIFAR-10 dataset. We hypothesize that the Content Gap accounts for a large portion of the performance drop when using synthetic images from DGM and propose strategies to better utilize them in downstream tasks. Extensive experiments on multiple datasets showcase that our method outperforms baselines on downstream classification tasks both in case of training on synthetic only (Synthetic-to-Real) and training on a mix of real and synthetic data (Data Augmentation), particularly in the data-scarce scenario.
Authors: Jiachen Li, Roberto Henschel, Vidit Goel, Marianna Ohanyan, Shant Navasardyan, Humphrey Shi
Conventional video matting outputs one alpha matte for all instances appearing in a video frame so that individual instances are not distinguished. While video instance segmentation provides time-consistent instance masks, results are unsatisfactory for matting applications, especially due to applied binarization. To remedy this deficiency, we propose Video Instance Matting~(VIM), that is, estimating alpha mattes of each instance at each frame of a video sequence. To tackle this challenging problem, we present MSG-VIM, a Mask Sequence Guided Video Instance Matting neural network, as a novel baseline model for VIM. MSG-VIM leverages a mixture of mask augmentations to make predictions robust to inaccurate and inconsistent mask guidance. It incorporates temporal mask and temporal feature guidance to improve the temporal consistency of alpha matte predictions. Furthermore, we build a new benchmark for VIM, called VIM50, which comprises 50 video clips with multiple human instances as foreground objects. To evaluate performances on the VIM task, we introduce a suitable metric called Video Instance-aware Matting Quality~(VIMQ). Our proposed model MSG-VIM sets a strong baseline on the VIM50 benchmark and outperforms existing methods by a large margin. The project is open-sourced at https://github.com/SHI-Labs/VIM.