Authors: Kazuki Hayashi, Yusuke Sakai, Hidetaka Kamigaito, Katsuhiko Hayashi, Taro Watanabe
Abstract: Large-scale vision-language models (LVLMs) output text from images and instructions, demonstrating advanced capabilities in text generation and comprehension. However, it has not been clarified to what extent LVLMs understand the knowledge necessary for explaining images, the complex relationships between various pieces of knowledge, and how they integrate these understandings into their explanations. To address this issue, we propose a new task: the artwork explanation generation task, along with its evaluation dataset and metric for quantitatively assessing the understanding and utilization of knowledge about artworks. This task is apt for image description based on the premise that LVLMs are expected to have pre-existing knowledge of artworks, which are often subjects of wide recognition and documented information. It consists of two parts: generating explanations from both images and titles of artworks, and generating explanations using only images, thus evaluating the LVLMs' language-based and vision-based knowledge. Alongside, we release a training dataset for LVLMs to learn explanations that incorporate knowledge about artworks. Our findings indicate that LVLMs not only struggle with integrating language and visual information but also exhibit a more pronounced limitation in acquiring knowledge from images alone. The datasets (ExpArt=Explain Artworks) are available at https://huggingface.co/datasets/naist-nlp/ExpArt.
Authors: Yu Wang, Mengying Xing
Abstract: Interest is increasing among political scientists in leveraging the extensive information available in images. However, the challenge of interpreting these images lies in the need for specialized knowledge in computer vision and access to specialized hardware. As a result, image analysis has been limited to a relatively small group within the political science community. This landscape could potentially change thanks to the rise of large language models (LLMs). This paper aims to raise awareness of the feasibility of using Gemini for image content analysis. A retrospective analysis was conducted on a corpus of 688 images. Content reports were elicited from Gemini for each image and then manually evaluated by the authors. We find that Gemini is highly accurate in performing object detection, which is arguably the most common and fundamental task in image analysis for political scientists. Equally important, we show that it is easy to implement as the entire command consists of a single prompt in natural language; it is fast to run and should meet the time budget of most researchers; and it is free to use and does not require any specialized hardware. In addition, we illustrate how political scientists can leverage Gemini for other image understanding tasks, including face identification, sentiment analysis, and caption generation. Our findings suggest that Gemini and other similar LLMs have the potential to drastically stimulate and accelerate image research in political science and social sciences more broadly.
Authors: Matthew Danish, SM Labib, Britta Ricker, Marco Helbich
Abstract: Street View-level Imagery (SVI) is a valuable data source for studies (e.g., environmental assessments, green space identification or land cover classification). While commercial SVI is available, such providers commonly restrict copying or reuse in ways necessary for research. Open SVI datasets are readily available from less restrictive sources, such as Mapillary, but due to the heterogeneity of the images, these require substantial preprocessing, filtering, and careful quality checks. We present an efficient method for automated downloading, processing, cropping, and filtering open SVI, to be used in a survey of human perceptions of the streets portrayed in these images. We demonstrate our open-source reusable SVI preparation and smartphone-friendly perception-survey software with Amsterdam (Netherlands) as the case study. Using a citizen science approach, we collected from 331 people 22,637 ratings about their perceptions for various criteria. We have published our software in a public repository for future re-use and reproducibility.
Authors: Safouane El Ghazouali, Youssef Mhirit, Ali Oukhrid, Umberto Michelucci, Hichem Nouira
Abstract: In the realm of computer vision, the integration of advanced techniques into the processing of RGB-D camera inputs poses a significant challenge, given the inherent complexities arising from diverse environmental conditions and varying object appearances. Therefore, this paper introduces FusionVision, an exhaustive pipeline adapted for the robust 3D segmentation of objects in RGB-D imagery. Traditional computer vision systems face limitations in simultaneously capturing precise object boundaries and achieving high-precision object detection on depth map as they are mainly proposed for RGB cameras. To address this challenge, FusionVision adopts an integrated approach by merging state-of-the-art object detection techniques, with advanced instance segmentation methods. The integration of these components enables a holistic (unified analysis of information obtained from both color \textit{RGB} and depth \textit{D} channels) interpretation of RGB-D data, facilitating the extraction of comprehensive and accurate object information. The proposed FusionVision pipeline employs YOLO for identifying objects within the RGB image domain. Subsequently, FastSAM, an innovative semantic segmentation model, is applied to delineate object boundaries, yielding refined segmentation masks. The synergy between these components and their integration into 3D scene understanding ensures a cohesive fusion of object detection and segmentation, enhancing overall precision in 3D object segmentation. The code and pre-trained models are publicly available at https://github.com/safouaneelg/FusionVision/.
Authors: Mathias Viborg Andersen, Ross Greer, Andreas M{\o}gelmose, Mohan Trivedi
Abstract: Advanced Driver Assistance Systems (ADAS) in intelligent vehicles rely on accurate driver perception within the vehicle cabin, often leveraging a combination of sensing modalities. However, these modalities operate at varying rates, posing challenges for real-time, comprehensive driver state monitoring. This paper addresses the issue of missing data due to sensor frame rate mismatches, introducing a generative model approach to create synthetic yet realistic thermal imagery. We propose using conditional generative adversarial networks (cGANs), specifically comparing the pix2pix and CycleGAN architectures. Experimental results demonstrate that pix2pix outperforms CycleGAN, and utilizing multi-view input styles, especially stacked views, enhances the accuracy of thermal image generation. Moreover, the study evaluates the model's generalizability across different subjects, revealing the importance of individualized training for optimal performance. The findings suggest the potential of generative models in addressing missing frames, advancing driver state monitoring for intelligent vehicles, and underscoring the need for continued research in model generalization and customization.
Authors: Takahiko Furuya
Abstract: Following the successes in the fields of vision and language, self-supervised pretraining via masked autoencoding of 3D point set data, or Masked Point Modeling (MPM), has achieved state-of-the-art accuracy in various downstream tasks. However, current MPM methods lack a property essential for 3D point set analysis, namely, invariance against rotation of 3D objects/scenes. Existing MPM methods are thus not necessarily suitable for real-world applications where 3D point sets may have inconsistent orientations. This paper develops, for the first time, a rotation-invariant self-supervised pretraining framework for practical 3D point set analysis. The proposed algorithm, called MaskLRF, learns rotation-invariant and highly generalizable latent features via masked autoencoding of 3D points within Local Reference Frames (LRFs), which are not affected by rotation of 3D point sets. MaskLRF enhances the quality of latent features by integrating feature refinement using relative pose encoding and feature reconstruction using low-level but rich 3D geometry. The efficacy of MaskLRF is validated via extensive experiments on diverse downstream tasks including classification, segmentation, registration, and domain adaptation. I confirm that MaskLRF achieves new state-of-the-art accuracies in analyzing 3D point sets having inconsistent orientations. Code will be available at: https://github.com/takahikof/MaskLRF
Authors: Pengyu Yan, Mahesh Bhosale, Jay Lal, Bikhyat Adhikari, David Doermann
Abstract: Chart visualizations are essential for data interpretation and communication; however, most charts are only accessible in image format and lack the corresponding data tables and supplementary information, making it difficult to alter their appearance for different application scenarios. To eliminate the need for original underlying data and information to perform chart editing, we propose ChartReformer, a natural language-driven chart image editing solution that directly edits the charts from the input images with the given instruction prompts. The key in this method is that we allow the model to comprehend the chart and reason over the prompt to generate the corresponding underlying data table and visual attributes for new charts, enabling precise edits. Additionally, to generalize ChartReformer, we define and standardize various types of chart editing, covering style, layout, format, and data-centric edits. The experiments show promising results for the natural language-driven chart image editing.
Authors: Yu Jing, Tan Yujuan, Ren Ao, Liu Duo
Abstract: The prediction of optical flow for occluded points is still a difficult problem that has not yet been solved. Recent methods use self-attention to find relevant non-occluded points as references for estimating the optical flow of occluded points based on the assumption of self-similarity. However, they rely on visual features of a single image and weak constraints, which are not sufficient to constrain the trained network to focus on erroneous and weakly relevant reference points. We make full use of online occlusion recognition information to construct occlusion extended visual features and two strong constraints, allowing the network to learn to focus only on the most relevant references without requiring occlusion ground truth to participate in the training of the network. Our method adds very few network parameters to the original framework, making it very lightweight. Extensive experiments show that our model has the greatest cross-dataset generalization. Our method achieves much greater error reduction, 18.6%, 16.2%, and 20.1% for all points, non-occluded points, and occluded points respectively from the state-of-the-art GMA-base method, MATCHFlow(GMA), on Sintel Albedo pass. Furthermore, our model achieves state-of-the-art performance on the Sintel bench-marks, ranking \#1 among all published methods on Sintel clean pass. The code will be open-source.
Authors: Xin Liu, Jiamin Wu, Tianzhu Zhang
Abstract: Large pre-trained Vision-Language Models (VLMs), like CLIP, exhibit strong generalization ability to downstream tasks but struggle in few-shot scenarios. Existing prompting techniques primarily focus on global text and image representations, yet overlooking multi-modal attribute characteristics. This limitation hinders the model's ability to perceive fine-grained visual details and restricts its generalization ability to a broader range of unseen classes. To address this issue, we propose a Multi-modal Attribute Prompting method (MAP) by jointly exploring textual attribute prompting, visual attribute prompting, and attribute-level alignment. The proposed MAP enjoys several merits. First, we introduce learnable visual attribute prompts enhanced by textual attribute semantics to adaptively capture visual attributes for images from unknown categories, boosting fine-grained visual perception capabilities for CLIP. Second, the proposed attribute-level alignment complements the global alignment to enhance the robustness of cross-modal alignment for open-vocabulary objects. To our knowledge, this is the first work to establish cross-modal attribute-level alignment for CLIP-based few-shot adaptation. Extensive experimental results on 11 datasets demonstrate that our method performs favorably against state-of-the-art approaches.
Authors: Lei Li, Yuqi Wang, Runxin Xu, Peiyi Wang, Xiachong Feng, Lingpeng Kong, Qi Liu
Abstract: Large vision-language models (LVLMs), exemplified by GPT-4V, excel across diverse tasks involving concrete images from natural scenes. However, their ability to interpret abstract figures, such as geometry shapes and scientific plots, remains limited due to a scarcity of training datasets in scientific domains. To fill this gap, we introduce Multimodal ArXiv, consisting of ArXivCap and ArXivQA, for enhancing LVLMs scientific comprehension. ArXivCap is a figure-caption dataset comprising 6.4M images and 3.9M captions sourced from 572K ArXiv papers spanning various scientific domains. Drawing from ArXivCap, we introduce ArXivQA, a question-answering dataset generated by prompting GPT-4V based on scientific figures. ArXivQA greatly enhances LVLMs' mathematical reasoning capabilities, achieving a 10.4% absolute accuracy gain on a multimodal mathematical reasoning benchmark. Furthermore, employing ArXivCap, we devise four vision-to-text tasks for benchmarking LVLMs. Evaluation results with state-of-the-art LVLMs underscore their struggle with the nuanced semantics of academic figures, with domain-specific training yielding substantial performance gains. Our error analysis uncovers misinterpretations of visual context, recognition errors, and the production of overly simplified captions by current LVLMs, shedding light on future improvements.
Authors: Suizhi Huang, Shalayiding Sirejiding, Yuxiang Lu, Yue Ding, Leheng Liu, Hui Zhou, Hongtao Lu
Abstract: Object detection and semantic segmentation are pivotal components in biomedical image analysis. Current single-task networks exhibit promising outcomes in both detection and segmentation tasks. Multi-task networks have gained prominence due to their capability to simultaneously tackle segmentation and detection tasks, while also accelerating the segmentation inference. Nevertheless, recent multi-task networks confront distinct limitations such as the difficulty in striking a balance between accuracy and inference speed. Additionally, they often overlook the integration of cross-scale features, which is especially important for biomedical image analysis. In this study, we propose an efficient end-to-end multi-task network capable of concurrently performing object detection and semantic segmentation called YOLO-Med. Our model employs a backbone and a neck for multi-scale feature extraction, complemented by the inclusion of two task-specific decoders. A cross-scale task-interaction module is employed in order to facilitate information fusion between various tasks. Our model exhibits promising results in balancing accuracy and speed when evaluated on the Kvasir-seg dataset and a private biomedical image dataset.
Authors: Haowei Liu, Yaya Shi, Haiyang Xu, Chunfeng Yuan, Qinghao Ye, Chenliang Li, Ming Yan, Ji Zhang, Fei Huang, Bing Li, Weiming Hu
Abstract: In vision-language pre-training (VLP), masked image modeling (MIM) has recently been introduced for fine-grained cross-modal alignment. However, in most existing methods, the reconstruction targets for MIM lack high-level semantics, and text is not sufficiently involved in masked modeling. These two drawbacks limit the effect of MIM in facilitating cross-modal semantic alignment. In this work, we propose a semantics-enhanced cross-modal MIM framework (SemMIM) for vision-language representation learning. Specifically, to provide more semantically meaningful supervision for MIM, we propose a local semantics enhancing approach, which harvest high-level semantics from global image features via self-supervised agreement learning and transfer them to local patch encodings by sharing the encoding space. Moreover, to achieve deep involvement of text during the entire MIM process, we propose a text-guided masking strategy and devise an efficient way of injecting textual information in both masked modeling and reconstruction target acquisition. Experimental results validate that our method improves the effectiveness of the MIM task in facilitating cross-modal semantic alignment. Compared to previous VLP models with similar model size and data scale, our SemMIM model achieves state-of-the-art or competitive performance on multiple downstream vision-language tasks.
Authors: Han Lu, Siyu Sun, Yichen Xie, Liqing Zhang, Xiaokang Yang, Junchi Yan
Abstract: In the long-tailed recognition field, the Decoupled Training paradigm has demonstrated remarkable capabilities among various methods. This paradigm decouples the training process into separate representation learning and classifier re-training. Previous works have attempted to improve both stages simultaneously, making it difficult to isolate the effect of classifier re-training. Furthermore, recent empirical studies have demonstrated that simple regularization can yield strong feature representations, emphasizing the need to reassess existing classifier re-training methods. In this study, we revisit classifier re-training methods based on a unified feature representation and re-evaluate their performances. We propose a new metric called Logits Magnitude as a superior measure of model performance, replacing the commonly used Weight Norm. However, since it is hard to directly optimize the new metric during training, we introduce a suitable approximate invariant called Regularized Standard Deviation. Based on the two newly proposed metrics, we prove that reducing the absolute value of Logits Magnitude when it is nearly balanced can effectively decrease errors and disturbances during training, leading to better model performance. Motivated by these findings, we develop a simple logits retargeting approach (LORT) without the requirement of prior knowledge of the number of samples per class. LORT divides the original one-hot label into small true label probabilities and large negative label probabilities distributed across each class. Our method achieves state-of-the-art performance on various imbalanced datasets, including CIFAR100-LT, ImageNet-LT, and iNaturalist2018.
Authors: Artur Wysoczanski, Nabil Ettehadi, Soroush Arabshahi, Yifei Sun, Karen Hinkley Stukovsky, Karol E. Watson, MeiLan K. Han, Erin D Michos, Alejandro P. Comellas, Eric A. Hoffman, Andrew F. Laine, R. Graham Barr, Elsa D. Angelini
Abstract: Pulmonary emphysema, the progressive, irreversible loss of lung tissue, is conventionally categorized into three subtypes identifiable on pathology and on lung computed tomography (CT) images. Recent work has led to the unsupervised learning of ten spatially-informed lung texture patterns (sLTPs) on lung CT, representing distinct patterns of emphysematous lung parenchyma based on both textural appearance and spatial location within the lung, and which aggregate into 6 robust and reproducible CT Emphysema Subtypes (CTES). Existing methods for sLTP segmentation, however, are slow and highly sensitive to changes in CT acquisition protocol. In this work, we present a robust 3-D squeeze-and-excitation CNN for supervised classification of sLTPs and CTES on lung CT. Our results demonstrate that this model achieves accurate and reproducible sLTP segmentation on lung CTscans, across two independent cohorts and independently of scanner manufacturer and model.
Authors: Jiahao Hong, Jialong Zuo, Chuchu Han, Ruochen Zheng, Ming Tian, Changxin Gao, Nong Sang
Abstract: Recent unsupervised person re-identification (re-ID) methods achieve high performance by leveraging fine-grained local context. These methods are referred to as part-based methods. However, most part-based methods obtain local contexts through horizontal division, which suffer from misalignment due to various human poses. Additionally, the misalignment of semantic information in part features restricts the use of metric learning, thus affecting the effectiveness of part-based methods. The two issues mentioned above result in the under-utilization of part features in part-based methods. We introduce the Spatial Cascaded Clustering and Weighted Memory (SCWM) method to address these challenges. SCWM aims to parse and align more accurate local contexts for different human body parts while allowing the memory module to balance hard example mining and noise suppression. Specifically, we first analyze the foreground omissions and spatial confusions issues in the previous method. Then, we propose foreground and space corrections to enhance the completeness and reasonableness of the human parsing results. Next, we introduce a weighted memory and utilize two weighting strategies. These strategies address hard sample mining for global features and enhance noise resistance for part features, which enables better utilization of both global and part features. Extensive experiments on Market-1501 and MSMT17 validate the proposed method's effectiveness over many state-of-the-art methods.
Authors: Kirill Prokhorov, Alexandr A. Kalinin
Abstract: Acne, a prevalent skin condition, necessitates precise severity assessment for effective treatment. Acne severity grading typically involves lesion counting and global assessment. However, manual grading suffers from variability and inefficiency, highlighting the need for automated tools. Recently, label distribution learning (LDL) was proposed as an effective framework for acne image grading, but its effectiveness is hindered by severity scales that assign varying numbers of lesions to different severity grades. Addressing these limitations, we proposed to incorporate severity scale information into lesion counting by combining LDL with label smoothing, and to decouple if from global assessment. A novel weighting scheme in our approach adjusts the degree of label smoothing based on the severity grading scale. This method helped to effectively manage label uncertainty without compromising class distinctiveness. Applied to the benchmark ACNE04 dataset, our model demonstrated improved performance in automated acne grading, showcasing its potential in enhancing acne diagnostics. The source code is publicly available at http://github.com/openface-io/acne-lds.
Authors: Wei Chen, Zichen Miao, Qiang Qiu
Abstract: To address the high computational and parameter complexity associated with fine-tuning large pre-trained models, researchers have developed parameter-efficient methods, where only partial parameters are updated for downstream tasks. However, these works often overlook the distinct properties of convolutional kernels, which still remain essential elements in many large models, such as Stable Diffusion. In this study, we first introduce filter subspace by decomposing convolutional kernels within each network layer over a small set of filter subspace elements, referred to as filter atoms. We then fine-tune these models to extract task-specific representation by only adapting the filter atoms, a few hundred parameters typically. To potentially expand the parameter space for tuning, we further show a simple approach to generate an overcomplete filter subspace by recursively decomposing each filter atom over another set of filter atoms. The fine-tuning of filter atoms reshapes the filter subspace, enabling convolutional layers to adapt to diverse downstream tasks efficiently. Extensive experiments show that such a simple scheme surpasses previous tuning baselines for both discriminate and generative tasks. Our approach can potentially be complementary to many existing fine-tuning methods.
Authors: Rohan Sarkar, Avinash Kak
Abstract: In the context of pose-invariant object recognition and retrieval, we demonstrate that it is possible to achieve significant improvements in performance if both the category-based and the object-identity-based embeddings are learned simultaneously during training. In hindsight, that sounds intuitive because learning about the categories is more fundamental than learning about the individual objects that correspond to those categories. However, to the best of what we know, no prior work in pose-invariant learning has demonstrated this effect. This paper presents an attention-based dual-encoder architecture with specially designed loss functions that optimize the inter- and intra-class distances simultaneously in two different embedding spaces, one for the category embeddings and the other for the object-level embeddings. The loss functions we have proposed are pose-invariant ranking losses that are designed to minimize the intra-class distances and maximize the inter-class distances in the dual representation spaces. We demonstrate the power of our approach with three challenging multi-view datasets, ModelNet-40, ObjectPI, and FG3D. With our dual approach, for single-view object recognition, we outperform the previous best by 20.0% on ModelNet40, 2.0% on ObjectPI, and 46.5% on FG3D. On the other hand, for single-view object retrieval, we outperform the previous best by 33.7% on ModelNet40, 18.8% on ObjectPI, and 56.9% on FG3D.
Authors: Xi Liu, Ying Guo, Cheng Zhen, Tong Li, Yingying Ao, Pengfei Yan
Abstract: Listening head generation aims to synthesize a non-verbal responsive listener head by modeling the correlation between the speaker and the listener in dynamic conversion.The applications of listener agent generation in virtual interaction have promoted many works achieving the diverse and fine-grained motion generation. However, they can only manipulate motions through simple emotional labels, but cannot freely control the listener's motions. Since listener agents should have human-like attributes (e.g. identity, personality) which can be freely customized by users, this limits their realism. In this paper, we propose a user-friendly framework called CustomListener to realize the free-form text prior guided listener generation. To achieve speaker-listener coordination, we design a Static to Dynamic Portrait module (SDP), which interacts with speaker information to transform static text into dynamic portrait token with completion rhythm and amplitude information. To achieve coherence between segments, we design a Past Guided Generation Module (PGG) to maintain the consistency of customized listener attributes through the motion prior, and utilize a diffusion-based structure conditioned on the portrait token and the motion prior to realize the controllable generation. To train and evaluate our model, we have constructed two text-annotated listening head datasets based on ViCo and RealTalk, which provide text-video paired labels. Extensive experiments have verified the effectiveness of our model.
Authors: Chen Duan, Pei Fu, Shan Guo, Qianyi Jiang, Xiaoming Wei
Abstract: In recent years, text-image joint pre-training techniques have shown promising results in various tasks. However, in Optical Character Recognition (OCR) tasks, aligning text instances with their corresponding text regions in images poses a challenge, as it requires effective alignment between text and OCR-Text (referring to the text in images as OCR-Text to distinguish from the text in natural language) rather than a holistic understanding of the overall image content. In this paper, we propose a new pre-training method called OCR-Text Destylization Modeling (ODM) that transfers diverse styles of text found in images to a uniform style based on the text prompt. With ODM, we achieve better alignment between text and OCR-Text and enable pre-trained models to adapt to the complex and diverse styles of scene text detection and spotting tasks. Additionally, we have designed a new labeling generation method specifically for ODM and combined it with our proposed Text-Controller module to address the challenge of annotation costs in OCR tasks, allowing a larger amount of unlabeled data to participate in pre-training. Extensive experiments on multiple public datasets demonstrate that our method significantly improves performance and outperforms current pre-training methods in scene text detection and spotting tasks. Code is available at {https://github.com/PriNing/ODM}.
Authors: Xueyuan Xu, Fulin Wei, Tianyuan Jia, Li Zhuo, Feiping Nie, Xia Wu
Abstract: In the last decade, embedded multi-label feature selection methods, incorporating the search for feature subsets into model optimization, have attracted considerable attention in accurately evaluating the importance of features in multi-label classification tasks. Nevertheless, the state-of-the-art embedded multi-label feature selection algorithms based on least square regression usually cannot preserve sufficient discriminative information in multi-label data. To tackle the aforementioned challenge, a novel embedded multi-label feature selection method, termed global redundancy and relevance optimization in orthogonal regression (GRROOR), is proposed to facilitate the multi-label feature selection. The method employs orthogonal regression with feature weighting to retain sufficient statistical and structural information related to local label correlations of the multi-label data in the feature learning process. Additionally, both global feature redundancy and global label relevancy information have been considered in the orthogonal regression model, which could contribute to the search for discriminative and non-redundant feature subsets in the multi-label data. The cost function of GRROOR is an unbalanced orthogonal Procrustes problem on the Stiefel manifold. A simple yet effective scheme is utilized to obtain an optimal solution. Extensive experimental results on ten multi-label data sets demonstrate the effectiveness of GRROOR.
Authors: Qiang Meng, Xiao Wang, JiaBao Wang, Liujiang Yan, Ke Wang
Abstract: Despite its compactness and information integrity, the range view representation of LiDAR data rarely occurs as the first choice for 3D perception tasks. In this work, we further push the envelop of the range-view representation with a novel multi-task framework, achieving unprecedented 3D detection performances. Our proposed Small, Versatile, and Mighty (SVM) network utilizes a pure convolutional architecture to fully unleash the efficiency and multi-tasking potentials of the range view representation. To boost detection performances, we first propose a range-view specific Perspective Centric Label Assignment (PCLA) strategy, and a novel View Adaptive Regression (VAR) module to further refine hard-to-predict box properties. In addition, our framework seamlessly integrates semantic segmentation and panoptic segmentation tasks for the LiDAR point cloud, without extra modules. Among range-view-based methods, our model achieves new state-of-the-art detection performances on the Waymo Open Dataset. Especially, over 10 mAP improvement over convolutional counterparts can be obtained on the vehicle class. Our presented results for other tasks further reveal the multi-task capabilities of the proposed small but mighty framework.
Authors: Guo Junjie, Gao Chenqiang, Liu Fangcen, Meng Deyu
Abstract: Infrared-visible object detection aims to achieve robust even full-day object detection by fusing the complementary information of infrared and visible images. However, highly dynamically variable complementary characteristics and commonly existing modality misalignment make the fusion of complementary information difficult. In this paper, we propose a Dynamic Adaptive Multispectral Detection Transformer (DAMS-DETR) based on DETR to simultaneously address these two challenges. Specifically, we propose a Modality Competitive Query Selection strategy to provide useful prior information. This strategy can dynamically select basic salient modality feature representation for each object. To effectively mine the complementary information and adapt to misalignment situations, we propose a Multispectral Deformable Cross-attention module to adaptively sample and aggregate multi-semantic level features of infrared and visible images for each object. In addition, we further adopt the cascade structure of DETR to better mine complementary information. Experiments on four public datasets of different scenes demonstrate significant improvements compared to other state-of-the-art methods. The code will be released at https://github.com/gjj45/DAMS-DETR.
Authors: Yuxiang Lu, Shalayiding Sirejiding, Bayram Bayramli, Suizhi Huang, Yue Ding, Hongtao Lu
Abstract: The task-conditional model is a distinctive stream for efficient multi-task learning. Existing works encounter a critical limitation in learning task-agnostic and task-specific representations, primarily due to shortcomings in global context modeling arising from CNN-based architectures, as well as a deficiency in multi-scale feature interaction within the decoder. In this paper, we introduce a novel task-conditional framework called Task Indicating Transformer (TIT) to tackle this challenge. Our approach designs a Mix Task Adapter module within the transformer block, which incorporates a Task Indicating Matrix through matrix decomposition, thereby enhancing long-range dependency modeling and parameter-efficient feature adaptation by capturing intra- and inter-task features. Moreover, we propose a Task Gate Decoder module that harnesses a Task Indicating Vector and gating mechanism to facilitate adaptive multi-scale feature refinement guided by task embeddings. Experiments on two public multi-task dense prediction benchmarks, NYUD-v2 and PASCAL-Context, demonstrate that our approach surpasses state-of-the-art task-conditional methods.
Authors: Ruiqian Nai, Zixin Wen, Ji Li, Yuanzhi Li, Yang Gao
Abstract: In representation learning, a disentangled representation is highly desirable as it encodes generative factors of data in a separable and compact pattern. Researchers have advocated leveraging disentangled representations to complete downstream tasks with encouraging empirical evidence. This paper further investigates the necessity of disentangled representation in downstream applications. Specifically, we show that dimension-wise disentangled representations are unnecessary on a fundamental downstream task, abstract visual reasoning. We provide extensive empirical evidence against the necessity of disentanglement, covering multiple datasets, representation learning methods, and downstream network architectures. Furthermore, our findings suggest that the informativeness of representations is a better indicator of downstream performance than disentanglement. Finally, the positive correlation between informativeness and disentanglement explains the claimed usefulness of disentangled representations in previous works. The source code is available at https://github.com/Richard-coder-Nai/disentanglement-lib-necessity.git.
URLs: https://github.com/Richard-coder-Nai/disentanglement-lib-necessity.git.
Authors: Xiaqiang Tang, Weigao Sun, Siyuan Hu, Yiyang Sun, Yafeng Guo
Abstract: The multi-modality and stochastic characteristics of human behavior make motion prediction a highly challenging task, which is critical for autonomous driving. While deep learning approaches have demonstrated their great potential in this area, it still remains unsolved to establish a connection between multiple driving scenes (e.g., merging, roundabout, intersection) and the design of deep learning models. Current learning-based methods typically use one unified model to predict trajectories in different scenarios, which may result in sub-optimal results for one individual scene. To address this issue, we propose Multi-Scenes Network (aka. MS-Net), which is a multi-path sparse model trained by an evolutionary process. MS-Net selectively activates a subset of its parameters during the inference stage to produce prediction results for each scene. In the training stage, the motion prediction task under differentiated scenes is abstracted as a multi-task learning problem, an evolutionary algorithm is designed to encourage the network search of the optimal parameters for each scene while sharing common knowledge between different scenes. Our experiment results show that with substantially reduced parameters, MS-Net outperforms existing state-of-the-art methods on well-established pedestrian motion prediction datasets, e.g., ETH and UCY, and ranks the 2nd place on the INTERACTION challenge.
Authors: Zhiying Leng, Tolga Birdal, Xiaohui Liang, Federico Tombari
Abstract: 3D shape generation from text is a fundamental task in 3D representation learning. The text-shape pairs exhibit a hierarchical structure, where a general text like "chair" covers all 3D shapes of the chair, while more detailed prompts refer to more specific shapes. Furthermore, both text and 3D shapes are inherently hierarchical structures. However, existing Text2Shape methods, such as SDFusion, do not exploit that. In this work, we propose HyperSDFusion, a dual-branch diffusion model that generates 3D shapes from a given text. Since hyperbolic space is suitable for handling hierarchical data, we propose to learn the hierarchical representations of text and 3D shapes in hyperbolic space. First, we introduce a hyperbolic text-image encoder to learn the sequential and multi-modal hierarchical features of text in hyperbolic space. In addition, we design a hyperbolic text-graph convolution module to learn the hierarchical features of text in hyperbolic space. In order to fully utilize these text features, we introduce a dual-branch structure to embed text features in 3D feature space. At last, to endow the generated 3D shapes with a hierarchical structure, we devise a hyperbolic hierarchical loss. Our method is the first to explore the hyperbolic hierarchical representation for text-to-shape generation. Experimental results on the existing text-to-shape paired dataset, Text2Shape, achieved state-of-the-art results.
Authors: Huan Ma, Yan Zhu, Changqing Zhang, Peilin Zhao, Baoyuan Wu, Long-Kai Huang, Qinghua Hu, Bingzhe Wu
Abstract: Vision-language foundation models have exhibited remarkable success across a multitude of downstream tasks due to their scalability on extensive image-text paired datasets. However, these models display significant limitations when applied to long-tail tasks, such as fine-grained image classification, as a result of "decision shortcuts" that hinders their generalization capabilities. In this work, we find that the CLIP model possesses a rich set of features, encompassing both \textit{desired invariant causal features} and \textit{undesired decision shortcuts}. Moreover, the underperformance of CLIP on downstream tasks originates from its inability to effectively utilize pre-trained features in accordance with specific task requirements. To address this challenge, this paper introduces a test-time prompt tuning paradigm that optimizes a learnable prompt, thereby compelling the model to exploit genuine causal invariant features while disregarding decision shortcuts during the inference phase. The proposed method effectively alleviates excessive dependence on potentially misleading, task-irrelevant contextual information, while concurrently emphasizing critical, task-related visual cues. We conduct comparative analysis of the proposed method against various approaches which validates its effectiveness.
Authors: Athanasios Tragakis, Qianying Liu, Chaitanya Kaul, Swalpa Kumar Roy, Hang Dai, Fani Deligianni, Roderick Murray-Smith, Daniele Faccio
Abstract: We propose a novel transformer-style architecture called Global-Local Filter Network (GLFNet) for medical image segmentation and demonstrate its state-of-the-art performance. We replace the self-attention mechanism with a combination of global-local filter blocks to optimize model efficiency. The global filters extract features from the whole feature map whereas the local filters are being adaptively created as 4x4 patches of the same feature map and add restricted scale information. In particular, the feature extraction takes place in the frequency domain rather than the commonly used spatial (image) domain to facilitate faster computations. The fusion of information from both spatial and frequency spaces creates an efficient model with regards to complexity, required data and performance. We test GLFNet on three benchmark datasets achieving state-of-the-art performance on all of them while being almost twice as efficient in terms of GFLOP operations.
Authors: Zhenpeng Huang, Chao Li, Hao Chen, Yongjian Deng, Yifeng Geng, Limin Wang
Abstract: In this paper, we present a new data-efficient voxel-based self-supervised learning method for event cameras. Our pre-training overcomes the limitations of previous methods, which either sacrifice temporal information by converting event sequences into 2D images for utilizing pre-trained image models or directly employ paired image data for knowledge distillation to enhance the learning of event streams. In order to make our pre-training data-efficient, we first design a semantic-uniform masking method to address the learning imbalance caused by the varying reconstruction difficulties of different regions in non-uniform data when using random masking. In addition, we ease the traditional hybrid masked modeling process by explicitly decomposing it into two branches, namely local spatio-temporal reconstruction and global semantic reconstruction to encourage the encoder to capture local correlations and global semantics, respectively. This decomposition allows our selfsupervised learning method to converge faster with minimal pre-training data. Compared to previous approaches, our self-supervised learning method does not rely on paired RGB images, yet enables simultaneous exploration of spatial and temporal cues in multiple scales. It exhibits excellent generalization performance and demonstrates significant improvements across various tasks with fewer parameters and lower computational costs.
Authors: Zhaorun Chen, Zhuokai Zhao, Hongyin Luo, Huaxiu Yao, Bo Li, Jiawei Zhou
Abstract: While large vision-language models (LVLMs) have demonstrated impressive capabilities in interpreting multi-modal contexts, they invariably suffer from object hallucinations (OH). We introduce HALC, a novel decoding algorithm designed to mitigate OH in LVLMs. HALC leverages distinct fine-grained optimal visual information in vision-language tasks and operates on both local and global contexts simultaneously. Specifically, HALC integrates a robust auto-focal grounding mechanism (locally) to correct hallucinated tokens on the fly, and a specialized beam search algorithm (globally) to significantly reduce OH while preserving text generation quality. Additionally, HALC can be integrated into any LVLMs as a plug-and-play module without extra training. Extensive experimental studies demonstrate the effectiveness of HALC in reducing OH, outperforming state-of-the-arts across four benchmarks.
Authors: Chengze Ye, Linda-Sophie Schneider, Yipeng Sun, Andreas Maier
Abstract: This study presents a novel approach for reconstructing cone beam computed tomography (CBCT) for specific orbits using known operator learning. Unlike traditional methods, this technique employs a filtered backprojection type (FBP-type) algorithm, which integrates a unique, adaptive filtering process. This process involves a series of operations, including weightings, differentiations, the 2D Radon transform, and backprojection. The filter is designed for a specific orbit geometry and is obtained using a data-driven approach based on deep learning. The approach efficiently learns and optimizes the orbit-related component of the filter. The method has demonstrated its ability through experimentation by successfully learning parameters from circular orbit projection data. Subsequently, the optimized parameters are used to reconstruct images, resulting in outcomes that closely resemble the analytical solution. This demonstrates the potential of the method to learn appropriate parameters from any specific orbit projection data and achieve reconstruction. The algorithm has demonstrated improvement, particularly in enhancing reconstruction speed and reducing memory usage for handling specific orbit reconstruction.
Authors: Jianwu Fang, Lei-lei Li, Junfei Zhou, Junbin Xiao, Hongkai Yu, Chen Lv, Jianru Xue, Tat-Seng Chua
Abstract: We present MM-AU, a novel dataset for Multi-Modal Accident video Understanding. MM-AU contains 11,727 in-the-wild ego-view accident videos, each with temporally aligned text descriptions. We annotate over 2.23 million object boxes and 58,650 pairs of video-based accident reasons, covering 58 accident categories. MM-AU supports various accident understanding tasks, particularly multimodal video diffusion to understand accident cause-effect chains for safe driving. With MM-AU, we present an Abductive accident Video understanding framework for Safe Driving perception (AdVersa-SD). AdVersa-SD performs video diffusion via an Object-Centric Video Diffusion (OAVD) method which is driven by an abductive CLIP model. This model involves a contrastive interaction loss to learn the pair co-occurrence of normal, near-accident, accident frames with the corresponding text descriptions, such as accident reasons, prevention advice, and accident categories. OAVD enforces the causal region learning while fixing the content of the original frame background in video generation, to find the dominant cause-effect chain for certain accidents. Extensive experiments verify the abductive ability of AdVersa-SD and the superiority of OAVD against the state-of-the-art diffusion models. Additionally, we provide careful benchmark evaluations for object detection and accident reason answering since AdVersa-SD relies on precise object and accident reason information.
Authors: Goirik Chakrabarty, Aditya Chandrasekar, Ramya Hebbalaguppe, Prathosh AP
Abstract: Recent developments in the field of diffusion models have demonstrated an exceptional capacity to generate high-quality prompt-conditioned image edits. Nevertheless, previous approaches have primarily relied on textual prompts for image editing, which tend to be less effective when making precise edits to specific objects or fine-grained regions within a scene containing single/multiple objects. We introduce a novel framework for zero-shot localized multi-object editing through a multi-diffusion process to overcome this challenge. This framework empowers users to perform various operations on objects within an image, such as adding, replacing, or editing $\textbf{many}$ objects in a complex scene $\textbf{in one pass}$. Our approach leverages foreground masks and corresponding simple text prompts that exert localized influences on the target regions resulting in high-fidelity image editing. A combination of cross-attention and background preservation losses within the latent space ensures that the characteristics of the object being edited are preserved while simultaneously achieving a high-quality, seamless reconstruction of the background with fewer artifacts compared to the current methods. We also curate and release a dataset dedicated to multi-object editing, named $\texttt{LoMOE}$-Bench. Our experiments against existing state-of-the-art methods demonstrate the improved effectiveness of our approach in terms of both image editing quality and inference speed.
Authors: Shumpei Takezaki, Seiichi Uchida
Abstract: Diffusion models have recently been used for medical image generation because of their high image quality. In this study, we focus on generating medical images with ordinal classes, which have ordinal relationships, such as severity levels. We propose an Ordinal Diffusion Model (ODM) that controls the ordinal relationships of the estimated noise images among the classes. Our model was evaluated experimentally by generating retinal and endoscopic images of multiple severity classes. ODM achieved higher performance than conventional generative models by generating realistic images, especially in high-severity classes with fewer training samples.
Authors: Yang Zhou, Zichong Chen, Hui Huang
Abstract: This paper addresses the complex issue of one-shot face stylization, focusing on the simultaneous consideration of appearance and structure, where previous methods have fallen short. We explore deformation-aware face stylization that diverges from traditional single-image style reference, opting for a real-style image pair instead. The cornerstone of our method is the utilization of a self-supervised vision transformer, specifically DINO-ViT, to establish a robust and consistent facial structure representation across both real and style domains. Our stylization process begins by adapting the StyleGAN generator to be deformation-aware through the integration of spatial transformers (STN). We then introduce two innovative constraints for generator fine-tuning under the guidance of DINO semantics: i) a directional deformation loss that regulates directional vectors in DINO space, and ii) a relative structural consistency constraint based on DINO token self-similarities, ensuring diverse generation. Additionally, style-mixing is employed to align the color generation with the reference, minimizing inconsistent correspondences. This framework delivers enhanced deformability for general one-shot face stylization, achieving notable efficiency with a fine-tuning duration of approximately 10 minutes. Extensive qualitative and quantitative comparisons demonstrate our superiority over state-of-the-art one-shot face stylization methods. Code is available at \url{https://github.com/zichongc/DoesFS}.
Authors: Wenjie Xuan, Yufei Xu, Shanshan Zhao, Chaoyue Wang, Juhua Liu, Bo Du, Dacheng Tao
Abstract: ControlNet excels at creating content that closely matches precise contours in user-provided masks. However, when these masks contain noise, as a frequent occurrence with non-expert users, the output would include unwanted artifacts. This paper first highlights the crucial role of controlling the impact of these inexplicit masks with diverse deterioration levels through in-depth analysis. Subsequently, to enhance controllability with inexplicit masks, an advanced Shape-aware ControlNet consisting of a deterioration estimator and a shape-prior modulation block is devised. The deterioration estimator assesses the deterioration factor of the provided masks. Then this factor is utilized in the modulation block to adaptively modulate the model's contour-following ability, which helps it dismiss the noise part in the inexplicit masks. Extensive experiments prove its effectiveness in encouraging ControlNet to interpret inaccurate spatial conditions robustly rather than blindly following the given contours. We showcase application scenarios like modifying shape priors and composable shape-controllable generation. Codes are soon available.
Authors: Yuanxin Liu, Shicheng Li, Yi Liu, Yuxiang Wang, Shuhuai Ren, Lei Li, Sishuo Chen, Xu Sun, Lu Hou
Abstract: Recently, there is a surge in interest surrounding video large language models (Video LLMs). However, existing benchmarks fail to provide a comprehensive feedback on the temporal perception ability of Video LLMs. On the one hand, most of them are unable to distinguish between different temporal aspects (e.g., speed, direction) and thus cannot reflect the nuanced performance on these specific aspects. On the other hand, they are limited in the diversity of task formats (e.g., only multi-choice QA), which hinders the understanding of how temporal perception performance may vary across different types of tasks. Motivated by these two problems, we propose the \textbf{TempCompass} benchmark, which introduces a diversity of temporal aspects and task formats. To collect high-quality test data, we devise two novel strategies: (1) In video collection, we construct conflicting videos that share the same static content but differ in a specific temporal aspect, which prevents Video LLMs from leveraging single-frame bias or language priors. (2) To collect the task instructions, we propose a paradigm where humans first annotate meta-information for a video and then an LLM generates the instruction. We also design an LLM-based approach to automatically and accurately evaluate the responses from Video LLMs. Based on TempCompass, we comprehensively evaluate 8 state-of-the-art (SOTA) Video LLMs and 3 Image LLMs, and reveal the discerning fact that these models exhibit notably poor temporal perception ability. Our data will be available at \url{https://github.com/llyx97/TempCompass}.
Authors: Mengqi Huang, Zhendong Mao, Mingcong Liu, Qian He, Yongdong Zhang
Abstract: Text-to-image customization, which aims to synthesize text-driven images for the given subjects, has recently revolutionized content creation. Existing works follow the pseudo-word paradigm, i.e., represent the given subjects as pseudo-words and then compose them with the given text. However, the inherent entangled influence scope of pseudo-words with the given text results in a dual-optimum paradox, i.e., the similarity of the given subjects and the controllability of the given text could not be optimal simultaneously. We present RealCustom that, for the first time, disentangles similarity from controllability by precisely limiting subject influence to relevant parts only, achieved by gradually narrowing real text word from its general connotation to the specific subject and using its cross-attention to distinguish relevance. Specifically, RealCustom introduces a novel "train-inference" decoupled framework: (1) during training, RealCustom learns general alignment between visual conditions to original textual conditions by a novel adaptive scoring module to adaptively modulate influence quantity; (2) during inference, a novel adaptive mask guidance strategy is proposed to iteratively update the influence scope and influence quantity of the given subjects to gradually narrow the generation of the real text word. Comprehensive experiments demonstrate the superior real-time customization ability of RealCustom in the open domain, achieving both unprecedented similarity of the given subjects and controllability of the given text for the first time. The project page is https://corleone-huang.github.io/realcustom/.
Authors: Xianqi Wang, Gangwei Xu, Hao Jia, Xin Yang
Abstract: Stereo matching methods based on iterative optimization, like RAFT-Stereo and IGEV-Stereo, have evolved into a cornerstone in the field of stereo matching. However, these methods struggle to simultaneously capture high-frequency information in edges and low-frequency information in smooth regions due to the fixed receptive field. As a result, they tend to lose details, blur edges, and produce false matches in textureless areas. In this paper, we propose Selective Recurrent Unit (SRU), a novel iterative update operator for stereo matching. The SRU module can adaptively fuse hidden disparity information at multiple frequencies for edge and smooth regions. To perform adaptive fusion, we introduce a new Contextual Spatial Attention (CSA) module to generate attention maps as fusion weights. The SRU empowers the network to aggregate hidden disparity information across multiple frequencies, mitigating the risk of vital hidden disparity information loss during iterative processes. To verify SRU's universality, we apply it to representative iterative stereo matching methods, collectively referred to as Selective-Stereo. Our Selective-Stereo ranks $1^{st}$ on KITTI 2012, KITTI 2015, ETH3D, and Middlebury leaderboards among all published methods. Code is available at https://github.com/Windsrain/Selective-Stereo.
Authors: Quentin Garrido, Mahmoud Assran, Nicolas Ballas, Adrien Bardes, Laurent Najman, Yann LeCun
Abstract: Joint-Embedding Predictive Architecture (JEPA) has emerged as a promising self-supervised approach that learns by leveraging a world model. While previously limited to predicting missing parts of an input, we explore how to generalize the JEPA prediction task to a broader set of corruptions. We introduce Image World Models, an approach that goes beyond masked image modeling and learns to predict the effect of global photometric transformations in latent space. We study the recipe of learning performant IWMs and show that it relies on three key aspects: conditioning, prediction difficulty, and capacity. Additionally, we show that the predictive world model learned by IWM can be adapted through finetuning to solve diverse tasks; a fine-tuned IWM world model matches or surpasses the performance of previous self-supervised methods. Finally, we show that learning with an IWM allows one to control the abstraction level of the learned representations, learning invariant representations such as contrastive methods, or equivariant representations such as masked image modelling.
Authors: Xiangxiang Chu, Jianlin Su, Bo Zhang, Chunhua Shen
Abstract: Large language models are built on top of a transformer-based architecture to process textual inputs. For example, the LLaMA stands out among many open-source implementations. Can the same transformer be used to process 2D images? In this paper, we answer this question by unveiling a LLaMA-like vision transformer in plain and pyramid forms, termed VisionLLaMA, which is tailored for this purpose. VisionLLaMA is a unified and generic modelling framework for solving most vision tasks. We extensively evaluate its effectiveness using typical pre-training paradigms in a good portion of downstream tasks of image perception and especially image generation. In many cases, VisionLLaMA have exhibited substantial gains over the previous state-of-the-art vision transformers. We believe that VisionLLaMA can serve as a strong new baseline model for vision generation and understanding. Our code will be released at https://github.com/Meituan-AutoML/VisionLLaMA.
Authors: Yuting Li, Yingyi Chen, Xuanlong Yu, Dexiong Chen, Xi Shen
Abstract: In this paper, we revisit techniques for uncertainty estimation within deep neural networks and consolidate a suite of techniques to enhance their reliability. Our investigation reveals that an integrated application of diverse techniques--spanning model regularization, classifier and optimization--substantially improves the accuracy of uncertainty predictions in image classification tasks. The synergistic effect of these techniques culminates in our novel SURE approach. We rigorously evaluate SURE against the benchmark of failure prediction, a critical testbed for uncertainty estimation efficacy. Our results showcase a consistently better performance than models that individually deploy each technique, across various datasets and model architectures. When applied to real-world challenges, such as data corruption, label noise, and long-tailed class distribution, SURE exhibits remarkable robustness, delivering results that are superior or on par with current state-of-the-art specialized methods. Particularly on Animal-10N and Food-101N for learning with noisy labels, SURE achieves state-of-the-art performance without any task-specific adjustments. This work not only sets a new benchmark for robust uncertainty estimation but also paves the way for its application in diverse, real-world scenarios where reliability is paramount. Our code is available at \url{https://yutingli0606.github.io/SURE/}.
Authors: Huaqing Yuan, Yi He, Peng Du, Lu Song
Abstract: Face images contain a wide variety of attribute information. In this paper, we propose a generalized framework for joint estimation of ordinal and nominal attributes based on information sharing. We tackle the correlation problem between heterogeneous attributes using hard parameter sharing of shallow features, and trade-off multiple loss functions by considering homoskedastic uncertainty for each attribute estimation task. This leads to optimal estimation of multiple attributes of the face and reduces the training cost of multitask learning. Experimental results on benchmarks with multiple face attributes show that the proposed approach has superior performance compared to state of the art. Finally, we discuss the bias issues arising from the proposed approach in face attribute estimation and validate its feasibility on edge systems.
Authors: Katherine Margaret Frances James, Karoline Heiwolt, Daniel James Sargent, Grzegorz Cielniak
Abstract: Automated phenotyping of plants for breeding and plant studies promises to provide quantitative metrics on plant traits at a previously unattainable observation frequency. Developers of tools for performing high-throughput phenotyping are, however, constrained by the availability of relevant datasets on which to perform validation. To this end, we present a spatio-temporal dataset of 3D point clouds of strawberry plants for two varieties, totalling 84 individual point clouds. We focus on the end use of such tools - the extraction of biologically relevant phenotypes - and demonstrate a phenotyping pipeline on the dataset. This comprises of the steps, including; segmentation, skeletonisation and tracking, and we detail how each stage facilitates the extraction of different phenotypes or provision of data insights. We particularly note that assessment is focused on the validation of phenotypes, extracted from the representations acquired at each step of the pipeline, rather than singularly focusing on assessing the representation itself. Therefore, where possible, we provide \textit{in silico} ground truth baselines for the phenotypes extracted at each step and introduce methodology for the quantitative assessment of skeletonisation and the length trait extracted thereof. This dataset contributes to the corpus of freely available agricultural/horticultural spatio-temporal data for the development of next-generation phenotyping tools, increasing the number of plant varieties available for research in this field and providing a basis for genuine comparison of new phenotyping methodology.
Authors: Yixiong Zou, Yicong Liu, Yiman Hu, Yuhua Li, Ruixuan Li
Abstract: Cross-domain few-shot learning (CDFSL) aims to acquire knowledge from limited training data in the target domain by leveraging prior knowledge transferred from source domains with abundant training samples. CDFSL faces challenges in transferring knowledge across dissimilar domains and fine-tuning models with limited training data. To address these challenges, we initially extend the analysis of loss landscapes from the parameter space to the representation space, which allows us to simultaneously interpret the transferring and fine-tuning difficulties of CDFSL models. We observe that sharp minima in the loss landscapes of the representation space result in representations that are hard to transfer and fine-tune. Moreover, existing flatness-based methods have limited generalization ability due to their short-range flatness. To enhance the transferability and facilitate fine-tuning, we introduce a simple yet effective approach to achieve long-range flattening of the minima in the loss landscape. This approach considers representations that are differently normalized as minima in the loss landscape and flattens the high-loss region in the middle by randomly sampling interpolated representations. We implement this method as a new normalization layer that replaces the original one in both CNNs and ViTs. This layer is simple and lightweight, introducing only a minimal number of additional parameters. Experimental results on 8 datasets demonstrate that our approach outperforms state-of-the-art methods in terms of average accuracy. Moreover, our method achieves performance improvements of up to 9\% compared to the current best approaches on individual datasets. Our code will be released.
Authors: Nikolas Adaloglou, Tim Kaiser, Felix Michels, Markus Kollmann
Abstract: We present a comprehensive experimental study on image-level conditioning for diffusion models using cluster assignments. We elucidate how individual components regarding image clustering impact image synthesis across three datasets. By combining recent advancements from image clustering and diffusion models, we show that, given the optimal cluster granularity with respect to image synthesis (visual groups), cluster-conditioning can achieve state-of-the-art FID (i.e. 1.67, 2.17 on CIFAR10 and CIFAR100 respectively), while attaining a strong training sample efficiency. Finally, we propose a novel method to derive an upper cluster bound that reduces the search space of the visual groups using solely feature-based clustering. Unlike existing approaches, we find no significant connection between clustering and cluster-conditional image generation. The code and cluster assignments will be released.
Authors: Musab Al-Ghadi, Joris Voerman, Souhail Bakkali, Micka\"el Coustaty, Nicolas Sidere, Xavier St-Georges
Abstract: The increasing use of digital technologies and mobile-based registration procedures highlights the vital role of personal identity documents (IDs) in verifying users and safeguarding sensitive information. However, the rise in counterfeit ID production poses a significant challenge, necessitating the development of reliable and efficient automated verification methods. This paper introduces IDTrust, a deep-learning framework for assessing the quality of IDs. IDTrust is a system that enhances the quality of identification documents by using a deep learning-based approach. This method eliminates the need for relying on original document patterns for quality checks and pre-processing steps for alignment. As a result, it offers significant improvements in terms of dataset applicability. By utilizing a bandpass filtering-based method, the system aims to effectively detect and differentiate ID quality. Comprehensive experiments on the MIDV-2020 and L3i-ID datasets identify optimal parameters, significantly improving discrimination performance and effectively distinguishing between original and scanned ID documents.
Authors: Ander Salaberria, Gorka Azkune, Oier Lopez de Lacalle, Aitor Soroa, Eneko Agirre, Frank Keller
Abstract: Existing work has observed that current text-to-image systems do not accurately reflect explicit spatial relations between objects such as 'left of' or 'below'. We hypothesize that this is because explicit spatial relations rarely appear in the image captions used to train these models. We propose an automatic method that, given existing images, generates synthetic captions that contain 14 explicit spatial relations. We introduce the Spatial Relation for Generation (SR4G) dataset, which contains 9.9 millions image-caption pairs for training, and more than 60 thousand captions for evaluation. In order to test generalization we also provide an 'unseen' split, where the set of objects in the train and test captions are disjoint. SR4G is the first dataset that can be used to spatially fine-tune text-to-image systems. We show that fine-tuning two different Stable Diffusion models (denoted as SD$_{SR4G}$) yields up to 9 points improvements in the VISOR metric. The improvement holds in the 'unseen' split, showing that SD$_{SR4G}$ is able to generalize to unseen objects. SD$_{SR4G}$ improves the state-of-the-art with fewer parameters, and avoids complex architectures. Our analysis shows that improvement is consistent for all relations. The dataset and the code will be publicly available.
Authors: Zhenwei He, Lei Zhang
Abstract: Object detection limits its recognizable categories during the training phase, in which it can not cover all objects of interest for users. To satisfy the practical necessity, the incremental learning ability of the detector becomes a critical factor for real-world applications. Unfortunately, neural networks unavoidably meet catastrophic forgetting problem when it is implemented on a new task. To this end, many incremental object detection models preserve the knowledge of previous tasks by replaying samples or distillation from previous models. However, they ignore an important factor that the performance of the model mostly depends on its feature. These models try to rouse the memory of the neural network with previous samples but not to prevent forgetting. To this end, in this paper, we propose an incremental causal object detection (ICOD) model by learning causal features, which can adapt to more tasks. Traditional object detection models, unavoidably depend on the data-bias or data-specific features to get the detection results, which can not adapt to the new task. When the model meets the requirements of incremental learning, the data-bias information is not beneficial to the new task, and the incremental learning may eliminate these features and lead to forgetting. To this end, our ICOD is introduced to learn the causal features, rather than the data-bias features when training the detector. Thus, when the model is implemented to a new task, the causal features of the old task can aid the incremental learning process to alleviate the catastrophic forgetting problem. We conduct our model on several experiments, which shows a causal feature without data-bias can make the model adapt to new tasks better. \keywords{Object detection, incremental learning, causal feature.
Authors: Zhaochong An, Guolei Sun, Yun Liu, Fayao Liu, Zongwei Wu, Dan Wang, Luc Van Gool, Serge Belongie
Abstract: This paper revisits few-shot 3D point cloud semantic segmentation (FS-PCS), with a focus on two significant issues in the state-of-the-art: foreground leakage and sparse point distribution. The former arises from non-uniform point sampling, allowing models to distinguish the density disparities between foreground and background for easier segmentation. The latter results from sampling only 2,048 points, limiting semantic information and deviating from the real-world practice. To address these issues, we introduce a standardized FS-PCS setting, upon which a new benchmark is built. Moreover, we propose a novel FS-PCS model. While previous methods are based on feature optimization by mainly refining support features to enhance prototypes, our method is based on correlation optimization, referred to as Correlation Optimization Segmentation (COSeg). Specifically, we compute Class-specific Multi-prototypical Correlation (CMC) for each query point, representing its correlations to category prototypes. Then, we propose the Hyper Correlation Augmentation (HCA) module to enhance CMC. Furthermore, tackling the inherent property of few-shot training to incur base susceptibility for models, we propose to learn non-parametric prototypes for the base classes during training. The learned base prototypes are used to calibrate correlations for the background class through a Base Prototypes Calibration (BPC) module. Experiments on popular datasets demonstrate the superiority of COSeg over existing methods. The code is available at: https://github.com/ZhaochongAn/COSeg
Authors: Zexin Feng, Na Zeng, Jiansheng Fang, Xingyue Wang, Xiaoxi Lu, Heng Meng, Jiang Liu
Abstract: Convolutional neural networks (CNNs) have long been the paradigm of choice for robust medical image processing (MIP). Therefore, it is crucial to effectively and efficiently deploy CNNs on devices with different computing capabilities to support computer-aided diagnosis. Many methods employ factorized convolutional layers to alleviate the burden of limited computational resources at the expense of expressiveness. To this end, given weak medical image-driven CNN model optimization, a Singular value equalization generalizer-induced Factorized Convolution (SFConv) is proposed to improve the expressive power of factorized convolutions in MIP models. We first decompose the weight matrix of convolutional filters into two low-rank matrices to achieve model reduction. Then minimize the KL divergence between the two low-rank weight matrices and the uniform distribution, thereby reducing the number of singular value directions with significant variance. Extensive experiments on fundus and OCTA datasets demonstrate that our SFConv yields competitive expressiveness over vanilla convolutions while reducing complexity.
Authors: Yuxi Liu, Wenhan Yang, Huihui Bai, Yunchao Wei, Yao Zhao
Abstract: Learned Image Compression (LIC) has shown remarkable progress in recent years. Existing works commonly employ CNN-based or self-attention-based modules as transform methods for compression. However, there is no prior research on neural transform that focuses on specific regions. In response, we introduce the class-agnostic segmentation masks (i.e. semantic masks without category labels) for extracting region-adaptive contextual information. Our proposed module, Region-Adaptive Transform, applies adaptive convolutions on different regions guided by the masks. Additionally, we introduce a plug-and-play module named Scale Affine Layer to incorporate rich contexts from various regions. While there have been prior image compression efforts that involve segmentation masks as additional intermediate inputs, our approach differs significantly from them. Our advantages lie in that, to avoid extra bitrate overhead, we treat these masks as privilege information, which is accessible during the model training stage but not required during the inference phase. To the best of our knowledge, we are the first to employ class-agnostic masks as privilege information and achieve superior performance in pixel-fidelity metrics, such as Peak Signal to Noise Ratio (PSNR). The experimental results demonstrate our improvement compared to previously well-performing methods, with about 8.2% bitrate saving compared to VTM-17.0. The code will be released at https://github.com/GityuxiLiu/Region-Adaptive-Transform-with-Segmentation-Prior-for-Image-Compression.
Authors: Yuhao Liu, Fang Liu, Zhanghan Ke, Nanxuan Zhao, Rynson W. H. Lau
Abstract: Diffusion models trained on large-scale datasets have achieved remarkable progress in image synthesis. However, due to the randomness in the diffusion process, they often struggle with handling diverse low-level tasks that require details preservation. To overcome this limitation, we present a new Diff-Plugin framework to enable a single pre-trained diffusion model to generate high-fidelity results across a variety of low-level tasks. Specifically, we first propose a lightweight Task-Plugin module with a dual branch design to provide task-specific priors, guiding the diffusion process in preserving image content. We then propose a Plugin-Selector that can automatically select different Task-Plugins based on the text instruction, allowing users to edit images by indicating multiple low-level tasks with natural language. We conduct extensive experiments on 8 low-level vision tasks. The results demonstrate the superiority of Diff-Plugin over existing methods, particularly in real-world scenarios. Our ablations further validate that Diff-Plugin is stable, schedulable, and supports robust training across different dataset sizes.
Authors: Lina Ruiz, Franklin Sierra-Jerez, Jair Ruiz, Fabio Martinez
Abstract: Colorectal cancer is the third most aggressive cancer worldwide. Polyps, as the main biomarker of the disease, are detected, localized, and characterized through colonoscopy procedures. Nonetheless, during the examination, up to 25% of polyps are missed, because of challenging conditions (camera movements, lighting changes), and the close similarity of polyps and intestinal folds. Besides, there is a remarked subjectivity and expert dependency to observe and detect abnormal regions along the intestinal tract. Currently, publicly available polyp datasets have allowed significant advances in computational strategies dedicated to characterizing non-parametric polyp shapes. These computational strategies have achieved remarkable scores of up to 90% in segmentation tasks. Nonetheless, these strategies operate on cropped and expert-selected frames that always observe polyps. In consequence, these computational approximations are far from clinical scenarios and real applications, where colonoscopies are redundant on intestinal background with high textural variability. In fact, the polyps typically represent less than 1% of total observations in a complete colonoscopy record. This work introduces COLON: the largest COlonoscopy LONg sequence dataset with around of 30 thousand polyp labeled frames and 400 thousand background frames. The dataset was collected from a total of 30 complete colonoscopies with polyps at different stages, variations in preparation procedures, and some cases the observation of surgical instrumentation. Additionally, 10 full intestinal background video control colonoscopies were integrated in order to achieve a robust polyp-background frame differentiation. The COLON dataset is open to the scientific community to bring new scenarios to propose computational tools dedicated to polyp detection and segmentation over long sequences, being closer to real colonoscopy scenarios.
Authors: Thomas Britton, Torri Jeske, David Lawrence, Kishansingh Rajput
Abstract: Hydra is a system which utilizes computer vision to perform near real time data quality management, initially developed for Hall-D in 2019. Since then, it has been deployed across all experimental halls at Jefferson Lab, with the CLAS12 collaboration in Hall-B being the first outside of GlueX to fully utilize Hydra. The system comprises back end processes that manage the models, their inferences, and the data flow. The front-end components, accessible via web pages, allow detector experts and shift crews to view and interact with the system. This talk will give an overview of the Hydra system as well as highlight significant developments in Hydra's feature set, acute challenges with operating Hydra in all halls, and lessons learned along the way.
Authors: Kangning Yin, Shihao Zou, Yuxuan Ge, Zheng Tian
Abstract: Information retrieval is an ever-evolving and crucial research domain. The substantial demand for high-quality human motion data especially in online acquirement has led to a surge in human motion research works. Prior works have mainly concentrated on dual-modality learning, such as text and motion tasks, but three-modality learning has been rarely explored. Intuitively, an extra introduced modality can enrich a model's application scenario, and more importantly, an adequate choice of the extra modality can also act as an intermediary and enhance the alignment between the other two disparate modalities. In this work, we introduce LAVIMO (LAnguage-VIdeo-MOtion alignment), a novel framework for three-modality learning integrating human-centric videos as an additional modality, thereby effectively bridging the gap between text and motion. Moreover, our approach leverages a specially designed attention mechanism to foster enhanced alignment and synergistic effects among text, video, and motion modalities. Empirically, our results on the HumanML3D and KIT-ML datasets show that LAVIMO achieves state-of-the-art performance in various motion-related cross-modal retrieval tasks, including text-to-motion, motion-to-text, video-to-motion and motion-to-video.
Authors: Gwangbin Bae, Andrew J. Davison
Abstract: Despite the growing demand for accurate surface normal estimation models, existing methods use general-purpose dense prediction models, adopting the same inductive biases as other tasks. In this paper, we discuss the inductive biases needed for surface normal estimation and propose to (1) utilize the per-pixel ray direction and (2) encode the relationship between neighboring surface normals by learning their relative rotation. The proposed method can generate crisp - yet, piecewise smooth - predictions for challenging in-the-wild images of arbitrary resolution and aspect ratio. Compared to a recent ViT-based state-of-the-art model, our method shows a stronger generalization ability, despite being trained on an orders of magnitude smaller dataset. The code is available at https://github.com/baegwangbin/DSINE.
Authors: Chuan Wen, Dinesh Jayaraman, Yang Gao
Abstract: Spatial relationships between objects represent key scene information for humans to understand and interact with the world. To study the capability of current computer vision systems to recognize physically grounded spatial relations, we start by proposing precise relation definitions that permit consistently annotating a benchmark dataset. Despite the apparent simplicity of this task relative to others in the recognition literature, we observe that existing approaches perform poorly on this benchmark. We propose new approaches exploiting the long-range attention capabilities of transformers for this task, and evaluating key design principles. We identify a simple "RelatiViT" architecture and demonstrate that it outperforms all current approaches. To our knowledge, this is the first method to convincingly outperform naive baselines on spatial relation prediction in in-the-wild settings. The code and datasets are available in \url{https://sites.google.com/view/spatial-relation}.
Authors: Tao Zhang, Xiangtai Li, Haobo Yuan, Shunping Ji, Shuicheng Yan
Abstract: In this work, for the first time, we demonstrate that Mamba-based point cloud methods can outperform point-based methods. Mamba exhibits strong global modeling capabilities and linear computational complexity, making it highly attractive for point cloud analysis. To enable more effective processing of 3-D point cloud data by Mamba, we propose a novel Consistent Traverse Serialization to convert point clouds into 1-D point sequences while ensuring that neighboring points in the sequence are also spatially adjacent. Consistent Traverse Serialization yields six variants by permuting the order of x, y, and z coordinates, and the synergistic use of these variants aids Mamba in comprehensively observing point cloud data. Furthermore, to assist Mamba in handling point sequences with different orders more effectively, we introduce point prompts to inform Mamba of the sequence's arrangement rules. Finally, we propose positional encoding based on spatial coordinate mapping to inject positional information into point cloud sequences better. Based on these improvements, we construct a point cloud network named Point Cloud Mamba, which combines local and global modeling. Point Cloud Mamba surpasses the SOTA point-based method PointNeXt and achieves new SOTA performance on the ScanObjectNN, ModelNet40, and ShapeNetPart datasets.
Authors: Karan Ahuja
Abstract: A long-standing vision in computer science has been to evolve computing devices into proactive assistants that enhance our productivity, health and wellness, and many other facets of our lives. User digitization is crucial in achieving this vision as it allows computers to intimately understand their users, capturing activity, pose, routine, and behavior. Today's consumer devices - like smartphones and smartwatches provide a glimpse of this potential, offering coarse digital representations of users with metrics such as step count, heart rate, and a handful of human activities like running and biking. Even these very low-dimensional representations are already bringing value to millions of people's lives, but there is significant potential for improvement. On the other end, professional, high-fidelity comprehensive user digitization systems exist. For example, motion capture suits and multi-camera rigs that digitize our full body and appearance, and scanning machines such as MRI capture our detailed anatomy. However, these carry significant user practicality burdens, such as financial, privacy, ergonomic, aesthetic, and instrumentation considerations, that preclude consumer use. In general, the higher the fidelity of capture, the lower the user's practicality. Most conventional approaches strike a balance between user practicality and digitization fidelity. My research aims to break this trend, developing sensing systems that increase user digitization fidelity to create new and powerful computing experiences while retaining or even improving user practicality and accessibility, allowing such technologies to have a societal impact. Armed with such knowledge, our future devices could offer longitudinal health tracking, more productive work environments, full body avatars in extended reality, and embodied telepresence experiences, to name just a few domains.
Authors: Aniket Tathe, Anand Kamble, Suyash Kumbharkar, Atharva Bhandare, Anirban C. Mitra
Abstract: This research addresses the challenge of training an ASR model for personalized voices with minimal data. Utilizing just 14 minutes of custom audio from a YouTube video, we employ Retrieval-Based Voice Conversion (RVC) to create a custom Common Voice 16.0 corpus. Subsequently, a Cross-lingual Self-supervised Representations (XLSR) Wav2Vec2 model is fine-tuned on this dataset. The developed web-based GUI efficiently transcribes and translates input Hindi videos. By integrating XLSR Wav2Vec2 and mBART, the system aligns the translated text with the video timeline, delivering an accessible solution for multilingual video content transcription and translation for personalized voice.
Authors: Chunlin Li, Ruofan Liang, Hanrui Fan, Zhengen Zhang, Sankeerth Durvasula, Nandita Vijaykumar
Abstract: We present a framework, DISORF, to enable online 3D reconstruction and visualization of scenes captured by resource-constrained mobile robots and edge devices. To address the limited compute capabilities of edge devices and potentially limited network availability, we design a framework that efficiently distributes computation between the edge device and remote server. We leverage on-device SLAM systems to generate posed keyframes and transmit them to remote servers that can perform high quality 3D reconstruction and visualization at runtime by leveraging NeRF models. We identify a key challenge with online NeRF training where naive image sampling strategies can lead to significant degradation in rendering quality. We propose a novel shifted exponential frame sampling method that addresses this challenge for online NeRF training. We demonstrate the effectiveness of our framework in enabling high-quality real-time reconstruction and visualization of unknown scenes as they are captured and streamed from cameras in mobile robots and edge devices.
Authors: Rukesh Prajapati, Amr S. El-Wakeel
Abstract: In contemporary rural healthcare settings, the principal challenge in diagnosing brain images is the scarcity of available data, given that most of the existing deep learning models demand extensive training data to optimize their performance, necessitating centralized processing methods that potentially compromise data privacy. This paper proposes a novel framework tailored for brain tissue segmentation in rural healthcare facilities. The framework employs a deep reinforcement learning (DRL) environment in tandem with a refinement model (RM) deployed locally at rural healthcare sites. The proposed DRL model has a reduced parameter count and practicality for implementation across distributed rural sites. To uphold data privacy and enhance model generalization without transgressing privacy constraints, we employ federated learning (FL) for cooperative model training. We demonstrate the efficacy of our approach by training the network with a limited data set and observing a substantial performance enhancement, mitigating inaccuracies and irregularities in segmentation across diverse sites. Remarkably, the DRL model attains an accuracy of up to 80%, surpassing the capabilities of conventional convolutional neural networks when confronted with data insufficiency. Incorporating our RM results in an additional accuracy improvement of at least 10%, while FL contributes to a further accuracy enhancement of up to 5%. Collectively, the framework achieves an average 92% accuracy rate within rural healthcare settings characterized by data constraints.
Authors: Wenjie Wei, Malu Zhang, Jilin Zhang, Ammar Belatreche, Jibin Wu, Zijing Xu, Xuerui Qiu, Hong Chen, Yang Yang, Haizhou Li
Abstract: Brain-inspired spiking neural networks (SNNs) have gained prominence in the field of neuromorphic computing owing to their low energy consumption during feedforward inference on neuromorphic hardware. However, it remains an open challenge how to effectively benefit from the sparse event-driven property of SNNs to minimize backpropagation learning costs. In this paper, we conduct a comprehensive examination of the existing event-driven learning algorithms, reveal their limitations, and propose novel solutions to overcome them. Specifically, we introduce two novel event-driven learning methods: the spike-timing-dependent event-driven (STD-ED) and membrane-potential-dependent event-driven (MPD-ED) algorithms. These proposed algorithms leverage precise neuronal spike timing and membrane potential, respectively, for effective learning. The two methods are extensively evaluated on static and neuromorphic datasets to confirm their superior performance. They outperform existing event-driven counterparts by up to 2.51% for STD-ED and 6.79% for MPD-ED on the CIFAR-100 dataset. In addition, we theoretically and experimentally validate the energy efficiency of our methods on neuromorphic hardware. On-chip learning experiments achieved a remarkable 30-fold reduction in energy consumption over time-step-based surrogate gradient methods. The demonstrated efficiency and efficacy of the proposed event-driven learning methods emphasize their potential to significantly advance the fields of neuromorphic computing, offering promising avenues for energy-efficiency applications.
Authors: C. Kosel, S. Michel, T. Seidel, M. Foerster
Abstract: Multimodal data analysis and validation based on streams from state-of-the-art sensor technology such as eye-tracking or emotion recognition using the Facial Action Coding System (FACTs) with deep learning allows educational researchers to study multifaceted learning and problem-solving processes and to improve educational experiences. This study aims to investigate the correlation between two continuous sensor streams, pupil diameter as an indicator of cognitive workload and FACTs with deep learning as an indicator of emotional arousal (RQ 1a), specifically for epochs of high, medium, and low arousal (RQ 1b). Furthermore, the time lag between emotional arousal and pupil diameter data will be analyzed (RQ 2). 28 participants worked on three cognitively demanding and emotionally engaging everyday moral dilemmas while eye-tracking and emotion recognition data were collected. The data were pre-processed in Phyton (synchronization, blink control, downsampling) and analyzed using correlation analysis and Granger causality tests. The results show negative and statistically significant correlations between the data streams for emotional arousal and pupil diameter. However, the correlation is negative and significant only for epochs of high arousal, while positive but non-significant relationships were found for epochs of medium or low arousal. The average time lag for the relationship between arousal and pupil diameter was 2.8 ms. In contrast to previous findings without a multimodal approach suggesting a positive correlation between the constructs, the results contribute to the state of research by highlighting the importance of multimodal data validation and research on convergent vagility. Future research should consider emotional regulation strategies and emotional valence.
Authors: Kibo Ote, Fumio Hashimoto, Yuya Onishi, Yasuomi Ouchi
Abstract: To converge the block iterative method in image reconstruction for positron emission tomography (PET), careful control of relaxation parameters is required, which is a challenging task. The automatic determination of relaxation parameters for list-mode reconstructions also remains challenging. Therefore, a different approach than controlling relaxation parameters would be desired by list-mode PET reconstruction. In this study, we propose a list-mode maximum likelihood Dykstra-like splitting PET reconstruction (LM-MLDS). LM-MLDS converges the list-mode block iterative method by adding the distance from an initial image as a penalty term into an objective function. LM-MLDS takes a two-step approach because its performance depends on the quality of the initial image. The first step uses a uniform image as the initial image, and then the second step uses a reconstructed image after one main iteration as the initial image. We evaluated LM-MLDS using simulation and clinical data. LM-MLDS provided a higher peak signal-to-noise ratio and suppressed an oscillation of tradeoff curves between noise and contrast than the other block iterative methods. In a clinical study, LM-MLDS removed the false hotspots at the edge of the axial field of view and improved the image quality of slices covering the top of the head to the cerebellum. LM-MLDS showed different noise properties than the other methods due to Gaussian denoising induced by the proximity operator. The list-mode proximal splitting PET reconstruction is useful not only for optimizing nondifferentiable functions such as total variation but also for converging block iterative methods without controlling relaxation parameters.
Authors: Utako Yamamoto, Hirohiko Imai, Kei Sano, Masayuki Ohzeki, Tetsuya Matsuda, Toshiyuki Tanaka
Abstract: The objective of our study is to observe dynamics of multiple substances in vivo with high temporal resolution from multi-spectral magnetic resonance spectroscopic imaging (MRSI) data. The multi-spectral MRSI can effectively separate spectral peaks of multiple substances and is useful to measure spatial distributions of substances. However it is difficult to measure time-varying substance distributions directly by ordinary full sampling because the measurement requires a significantly long time. In this study, we propose a novel method to reconstruct the spatio-temporal distributions of substances from randomly undersampled multi-spectral MRSI data on the basis of compressed sensing (CS) and the partially separable function model with base spectra of substances. In our method, we have employed spatio-temporal sparsity and temporal smoothness of the substance distributions as prior knowledge to perform CS. The effectiveness of our method has been evaluated using phantom data sets of glass tubes filled with glucose or lactate solution in increasing amounts over time and animal data sets of a tumor-bearing mouse to observe the metabolic dynamics involved in the Warburg effect in vivo. The reconstructed results are consistent with the expected behaviors, showing that our method can reconstruct the spatio-temporal distribution of substances with a temporal resolution of four seconds which is extremely short time scale compared with that of full sampling. Since this method utilizes only prior knowledge naturally assumed for the spatio-temporal distributions of substances and is independent of the number of the spectral and spatial dimensions or the acquisition sequence of MRSI, it is expected to contribute to revealing the underlying substance dynamics in MRSI data already acquired or to be acquired in the future.
Authors: Edward Whittaker, Ikuo Kitagishi
Abstract: Language Models (LMs) such as BERT, have been shown to perform well on the task of identifying Named Entities (NE) in text. A BERT LM is typically used as a classifier to classify individual tokens in the input text, or to classify spans of tokens, as belonging to one of a set of possible NE categories. In this paper, we hypothesise that decoder-only Large Language Models (LLMs) can also be used generatively to extract both the NE, as well as potentially recover the correct surface form of the NE, where any spelling errors that were present in the input text get automatically corrected. We fine-tune two BERT LMs as baselines, as well as eight open-source LLMs, on the task of producing NEs from text that was obtained by applying Optical Character Recognition (OCR) to images of Japanese shop receipts; in this work, we do not attempt to find or evaluate the location of NEs in the text. We show that the best fine-tuned LLM performs as well as, or slightly better than, the best fine-tuned BERT LM, although the differences are not significant. However, the best LLM is also shown to correct OCR errors in some cases, as initially hypothesised.
Authors: Yidong Zhao, Yi Zhang, Qian Tao
Abstract: Deep learning-based methods have achieved prestigious performance for magnetic resonance imaging (MRI) reconstruction, enabling fast imaging for many clinical applications. Previous methods employ convolutional networks to learn the image prior as the regularization term. In quantitative MRI, the physical model of nuclear magnetic resonance relaxometry is known, providing additional prior knowledge for image reconstruction. However, traditional reconstruction networks are limited to learning the spatial domain prior knowledge, ignoring the relaxometry prior. Therefore, we propose a relaxometry-guided quantitative MRI reconstruction framework to learn the spatial prior from data and the relaxometry prior from MRI physics. Additionally, we also evaluated the performance of two popular reconstruction backbones, namely, recurrent variational networks (RVN) and variational networks (VN) with U- Net. Experiments demonstrate that the proposed method achieves highly promising results in quantitative MRI reconstruction.
Authors: Martin J. Hetz, Carina Nogueira Garcia, Sarah Haggenm\"uller, Titus J. Brinker
Abstract: Clinical dermatology necessitates precision and innovation for efficient diagnosis and treatment of various skin conditions. This paper introduces the development of a cutting-edge hyperspectral dermatoscope (the Hyperscope) tailored for human skin analysis. We detail the requirements to such a device and the design considerations, from optical configurations to sensor selection, necessary to capture a wide spectral range with high fidelity. Preliminary results from 15 individuals and 160 recorded skin images demonstrate the potential of the Hyperscope in identifying and characterizing various skin conditions, offering a promising avenue for non-invasive skin evaluation and a platform for future research in dermatology-related hyperspectral imaging.
Authors: Leopold Hebert-Stevens, Gabriel Jimenez, Benoit Delatour, Lev Stimmer, Daniel Racoceanu
Abstract: This study utilizes graph theory and deep learning to assess variations in Alzheimer's disease (AD) neuropathologies, focusing on classic (cAD) and rapid (rpAD) progression forms. It analyses the distribution of amyloid plaques and tau tangles in postmortem brain tissues. Histopathological images are converted into tau-pathology-based graphs, and derived metrics are used for statistical analysis and in machine learning classifiers. These classifiers incorporate SHAP value explainability to differentiate between cAD and rpAD. Graph neural networks (GNNs) demonstrate greater efficiency than traditional CNN methods in analyzing this data, preserving spatial pathology context. Additionally, GNNs provide significant insights through explainable AI techniques. The analysis shows denser networks in rpAD and a distinctive impact on brain cortical layers: rpAD predominantly affects middle layers, whereas cAD influences both superficial and deep layers of the same cortical regions. These results suggest a unique neuropathological network organization for each AD variant.
Authors: Xianghong Fang, Jian Li, Qiang Sun, Benyou Wang
Abstract: Uniformity plays a crucial role in the assessment of learned representations, contributing to a deeper comprehension of self-supervised learning. The seminal work by \citet{Wang2020UnderstandingCR} introduced a uniformity metric that quantitatively measures the collapse degree of learned representations. Directly optimizing this metric together with alignment proves to be effective in preventing constant collapse. However, we present both theoretical and empirical evidence revealing that this metric lacks sensitivity to dimensional collapse, highlighting its limitations. To address this limitation and design a more effective uniformity metric, this paper identifies five fundamental properties, some of which the existing uniformity metric fails to meet. We subsequently introduce a novel uniformity metric that satisfies all of these desiderata and exhibits sensitivity to dimensional collapse. When applied as an auxiliary loss in various established self-supervised methods, our proposed uniformity metric consistently enhances their performance in downstream tasks.Our code was released at https://github.com/sunset-clouds/WassersteinUniformityMetric.
URLs: https://github.com/sunset-clouds/WassersteinUniformityMetric.
Authors: Jiaying Gong, Hoda Eldardiry
Abstract: The goal of few-shot relation extraction is to predict relations between name entities in a sentence when only a few labeled instances are available for training. Existing few-shot relation extraction methods focus on uni-modal information such as text only. This reduces performance when there are no clear contexts between the name entities described in text. We propose a multi-modal few-shot relation extraction model (MFS-HVE) that leverages both textual and visual semantic information to learn a multi-modal representation jointly. The MFS-HVE includes semantic feature extractors and multi-modal fusion components. The MFS-HVE semantic feature extractors are developed to extract both textual and visual features. The visual features include global image features and local object features within the image. The MFS-HVE multi-modal fusion unit integrates information from various modalities using image-guided attention, object-guided attention, and hybrid feature attention to fully capture the semantic interaction between visual regions of images and relevant texts. Extensive experiments conducted on two public datasets demonstrate that semantic visual information significantly improves the performance of few-shot relation prediction.
Authors: Peter Lorenz, Margret Keuper, Janis Keuper
Abstract: Convolutional neural networks (CNN) define the state-of-the-art solution on many perceptual tasks. However, current CNN approaches largely remain vulnerable against adversarial perturbations of the input that have been crafted specifically to fool the system while being quasi-imperceptible to the human eye. In recent years, various approaches have been proposed to defend CNNs against such attacks, for example by model hardening or by adding explicit defence mechanisms. Thereby, a small "detector" is included in the network and trained on the binary classification task of distinguishing genuine data from data containing adversarial perturbations. In this work, we propose a simple and light-weight detector, which leverages recent findings on the relation between networks' local intrinsic dimensionality (LID) and adversarial attacks. Based on a re-interpretation of the LID measure and several simple adaptations, we surpass the state-of-the-art on adversarial detection by a significant margin and reach almost perfect results in terms of F1-score for several networks and datasets. Sources available at: https://github.com/adverML/multiLID
Authors: Rim Rekik, Mathieu Marsot, Stefanie Wuhrer, Jean-S\'ebastien Franco, Anne-H\'el\`ene Olivier
Abstract: We present a data-driven framework for unsupervised human motion retargeting that animates a target subject with the motion of a source subject. Our method is correspondence-free, requiring neither spatial correspondences between the source and target shapes nor temporal correspondences between different frames of the source motion. This allows to animate a target shape with arbitrary sequences of humans in motion, possibly captured using 4D acquisition platforms or consumer devices. Our method unifies the advantages of two existing lines of work, namely skeletal motion retargeting, which leverages long-term temporal context, and surface-based retargeting, which preserves surface details, by combining a geometry-aware deformation model with a skeleton-aware motion transfer approach. This allows to take into account long-term temporal context while accounting for surface details. During inference, our method runs online, i.e. input can be processed in a serial way, and retargeting is performed in a single forward pass per frame. Experiments show that including long-term temporal context during training improves the method's accuracy for skeletal motion and detail preservation. Furthermore, our method generalizes to unobserved motions and body shapes. We demonstrate that our method achieves state-of-the-art results on two test datasets and that it can be used to animate human models with the output of a multi-view acquisition platform. Code is available at \url{https://gitlab.inria.fr/rrekikdi/human-motion-retargeting2023}.
URLs: https://gitlab.inria.fr/rrekikdi/human-motion-retargeting2023
Authors: Jiaqi Xu, Bo Liu, Yunkuo Chen, Mengli Cheng, Xing Shi
Abstract: Video-and-language understanding has a variety of applications in the industry, such as video question answering, text-video retrieval, and multi-label classification. Existing video-and-language understanding methods generally adopt heavy multi-modal encoders and feature fusion modules, which consume high computational costs. Specially, they have difficulty dealing with dense video frames or long text prevalent in industrial applications. This paper proposes MuLTI, a highly accurate and efficient video-and-language understanding model that achieves efficient and effective feature fusion and rapid adaptation to downstream tasks. Specifically, we design a Text-Guided MultiWay-Sampler based on adapt-pooling residual mapping and self-attention modules to sample long sequences and fuse multi-modal features, which reduces the computational costs and addresses performance degradation caused by previous samplers. Therefore, MuLTI can handle longer sequences with limited computational costs. Then, to further enhance the model's performance and fill in the lack of pretraining tasks in the video question answering, we propose a new pretraining task named Multiple Choice Modeling. This task bridges the gap between pretraining and downstream tasks and improves the model's ability to align video and text features. Benefiting from the efficient feature fusion module and the new pretraining task, MuLTI achieves state-of-the-art performance on multiple datasets. Implementation and pretrained models will be released.
Authors: Shuai Chen, Yash Bhalgat, Xinghui Li, Jiawang Bian, Kejie Li, Zirui Wang, Victor Adrian Prisacariu
Abstract: Absolute Pose Regression (APR) methods use deep neural networks to directly regress camera poses from RGB images. However, the predominant APR architectures only rely on 2D operations during inference, resulting in limited accuracy of pose estimation due to the lack of 3D geometry constraints or priors. In this work, we propose a test-time refinement pipeline that leverages implicit geometric constraints using a robust feature field to enhance the ability of APR methods to use 3D information during inference. We also introduce a novel Neural Feature Synthesizer (NeFeS) model, which encodes 3D geometric features during training and directly renders dense novel view features at test time to refine APR methods. To enhance the robustness of our model, we introduce a feature fusion module and a progressive training strategy. Our proposed method achieves state-of-the-art single-image APR accuracy on indoor and outdoor datasets.
Authors: Chenpeng Du, Qi Chen, Xie Chen, Kai Yu
Abstract: While recent research has made significant progress in speech-driven talking face generation, the quality of the generated video still lags behind that of real recordings. One reason for this is the use of handcrafted intermediate representations like facial landmarks and 3DMM coefficients, which are designed based on human knowledge and are insufficient to precisely describe facial movements. Additionally, these methods require an external pretrained model for extracting these representations, whose performance sets an upper bound on talking face generation. To address these limitations, we propose a novel method called DAE-Talker that leverages data-driven latent representations obtained from a diffusion autoencoder (DAE). DAE contains an image encoder that encodes an image into a latent vector and a DDIM image decoder that reconstructs the image from it. We train our DAE on talking face video frames and then extract their latent representations as the training target for a Conformer-based speech2latent model. This allows DAE-Talker to synthesize full video frames and produce natural head movements that align with the content of speech, rather than relying on a predetermined head pose from a template video. We also introduce pose modelling in speech2latent for pose controllability. Additionally, we propose a novel method for generating continuous video frames with the DDIM image decoder trained on individual frames, eliminating the need for modelling the joint distribution of consecutive frames directly. Our experiments show that DAE-Talker outperforms existing popular methods in lip-sync, video fidelity, and pose naturalness. We also conduct ablation studies to analyze the effectiveness of the proposed techniques and demonstrate the pose controllability of DAE-Talker.
Authors: Nikita Srivatsan, Sofia Samaniego, Omar Florez, Taylor Berg-Kirkpatrick
Abstract: In this work we present an approach for generating alternative text (or alt-text) descriptions for images shared on social media, specifically Twitter. More than just a special case of image captioning, alt-text is both more literally descriptive and context-specific. Also critically, images posted to Twitter are often accompanied by user-written text that despite not necessarily describing the image may provide useful context that if properly leveraged can be informative. We address this task with a multimodal model that conditions on both textual information from the associated social media post as well as visual signal from the image, and demonstrate that the utility of these two information sources stacks. We put forward a new dataset of 371k images paired with alt-text and tweets scraped from Twitter and evaluate on it across a variety of automated metrics as well as human evaluation. We show that our approach of conditioning on both tweet text and visual information significantly outperforms prior work, by more than 2x on BLEU@4.
Authors: Junyi Chen, Longteng Guo, Jia Sun, Shuai Shao, Zehuan Yuan, Liang Lin, Dongyu Zhang
Abstract: Building scalable vision-language models to learn from diverse, multimodal data remains an open challenge. In this paper, we introduce an Efficient Vision-languagE foundation model, namely EVE, which is one unified multimodal Transformer pre-trained solely by one unified pre-training task. Specifically, EVE encodes both vision and language within a shared Transformer network integrated with modality-aware sparse Mixture-of-Experts (MoE) modules, which capture modality-specific information by selectively switching to different experts. To unify pre-training tasks of vision and language, EVE performs masked signal modeling on image-text pairs to reconstruct masked signals, i.e., image pixels and text tokens, given visible signals. This simple yet effective pre-training objective accelerates training by 3.5x compared to the model pre-trained with Image-Text Contrastive and Image-Text Matching losses. Owing to the combination of the unified architecture and pre-training task, EVE is easy to scale up, enabling better downstream performance with fewer resources and faster training speed. Despite its simplicity, EVE achieves state-of-the-art performance on various vision-language downstream tasks, including visual question answering, visual reasoning, and image-text retrieval.
Authors: Markus K\"appeler, K\"ursat Petek, Niclas V\"odisch, Wolfram Burgard, Abhinav Valada
Abstract: Current state-of-the-art methods for panoptic segmentation require an immense amount of annotated training data that is both arduous and expensive to obtain posing a significant challenge for their widespread adoption. Concurrently, recent breakthroughs in visual representation learning have sparked a paradigm shift leading to the advent of large foundation models that can be trained with completely unlabeled images. In this work, we propose to leverage such task-agnostic image features to enable few-shot panoptic segmentation by presenting Segmenting Panoptic Information with Nearly 0 labels (SPINO). In detail, our method combines a DINOv2 backbone with lightweight network heads for semantic segmentation and boundary estimation. We show that our approach, albeit being trained with only ten annotated images, predicts high-quality pseudo-labels that can be used with any existing panoptic segmentation method. Notably, we demonstrate that SPINO achieves competitive results compared to fully supervised baselines while using less than 0.3% of the ground truth labels, paving the way for learning complex visual recognition tasks leveraging foundation models. To illustrate its general applicability, we further deploy SPINO on real-world robotic vision systems for both outdoor and indoor environments. To foster future research, we make the code and trained models publicly available at http://spino.cs.uni-freiburg.de.
Authors: Yuren Cong, Mengmeng Xu, Christian Simon, Shoufa Chen, Jiawei Ren, Yanping Xie, Juan-Manuel Perez-Rua, Bodo Rosenhahn, Tao Xiang, Sen He
Abstract: Text-to-video editing aims to edit the visual appearance of a source video conditional on textual prompts. A major challenge in this task is to ensure that all frames in the edited video are visually consistent. Most recent works apply advanced text-to-image diffusion models to this task by inflating 2D spatial attention in the U-Net into spatio-temporal attention. Although temporal context can be added through spatio-temporal attention, it may introduce some irrelevant information for each patch and therefore cause inconsistency in the edited video. In this paper, for the first time, we introduce optical flow into the attention module in the diffusion model's U-Net to address the inconsistency issue for text-to-video editing. Our method, FLATTEN, enforces the patches on the same flow path across different frames to attend to each other in the attention module, thus improving the visual consistency in the edited videos. Additionally, our method is training-free and can be seamlessly integrated into any diffusion-based text-to-video editing methods and improve their visual consistency. Experiment results on existing text-to-video editing benchmarks show that our proposed method achieves the new state-of-the-art performance. In particular, our method excels in maintaining the visual consistency in the edited videos.
Authors: Adam Schmidt, Omid Mohareri, Simon DiMaio, Michael C. Yip, Septimiu E. Salcudean
Abstract: As computer vision algorithms increase in capability, their applications in clinical systems will become more pervasive. These applications include: diagnostics, such as colonoscopy and bronchoscopy; guiding biopsies, minimally invasive interventions, and surgery; automating instrument motion; and providing image guidance using pre-operative scans. Many of these applications depend on the specific visual nature of medical scenes and require designing algorithms to perform in this environment. In this review, we provide an update to the field of camera-based tracking and scene mapping in surgery and diagnostics in medical computer vision. We begin with describing our review process, which results in a final list of 515 papers that we cover. We then give a high-level summary of the state of the art and provide relevant background for those who need tracking and mapping for their clinical applications. After which, we review datasets provided in the field and the clinical needs that motivate their design. Then, we delve into the algorithmic side, and summarize recent developments. This summary should be especially useful for algorithm designers and to those looking to understand the capability of off-the-shelf methods. We maintain focus on algorithms for deformable environments while also reviewing the essential building blocks in rigid tracking and mapping since there is a large amount of crossover in methods. With the field summarized, we discuss the current state of the tracking and mapping methods along with needs for future algorithms, needs for quantification, and the viability of clinical applications. We then provide some research directions and questions. We conclude that new methods need to be designed or combined to support clinical applications in deformable environments, and more focus needs to be put into collecting datasets for training and evaluation.
Authors: Chuanxin Song, Hanbo Wu, Xin Ma
Abstract: Variable scene layouts and coexisting objects across scenes make indoor scene recognition still a challenging task. Leveraging object information within scenes to enhance the distinguishability of feature representations has emerged as a key approach in this domain. Currently, most object-assisted methods use a separate branch to process object information, combining object and scene features heuristically. However, few of them pay attention to interpretably handle the hidden discriminative knowledge within object information. In this paper, we propose to leverage discriminative object knowledge to enhance scene feature representations. Initially, we capture the object-scene discriminative relationships from a probabilistic perspective, which are transformed into an Inter-Object Discriminative Prototype (IODP). Given the abundant prior knowledge from IODP, we subsequently construct a Discriminative Graph Network (DGN), in which pixel-level scene features are defined as nodes and the discriminative relationships between node features are encoded as edges. DGN aims to incorporate inter-object discriminative knowledge into the image representation through graph convolution and mapping operations (GCN). With the proposed IODP and DGN, we obtain state-of-the-art results on several widely used scene datasets, demonstrating the effectiveness of the proposed approach.
Authors: Zeyu Zhang, Xuyin Qi, Bowen Zhang, Biao Wu, Hien Le, Bora Jeong, Zhibin Liao, Yunxiang Liu, Johan Verjans, Minh-Son To, Richard Hartley
Abstract: Organ at risk (OAR) segmentation is a critical process in radiotherapy treatment planning such as head and neck tumors. Nevertheless, in clinical practice, radiation oncologists predominantly perform OAR segmentations manually on CT scans. This manual process is highly time-consuming and expensive, limiting the number of patients who can receive timely radiotherapy. Additionally, CT scans offer lower soft-tissue contrast compared to MRI. Despite MRI providing superior soft-tissue visualization, its time-consuming nature makes it infeasible for real-time treatment planning. To address these challenges, we propose a method called SegReg, which utilizes Elastic Symmetric Normalization for registering MRI to perform OAR segmentation. SegReg outperforms the CT-only baseline by 16.78% in mDSC and 18.77% in mIoU, showing that it effectively combines the geometric accuracy of CT with the superior soft-tissue contrast of MRI, making accurate automated OAR segmentation for clinical practice become possible. See project website https://steve-zeyu-zhang.github.io/SegReg
Authors: Yuxin Du, Fan Bai, Tiejun Huang, Bo Zhao
Abstract: Precise image segmentation provides clinical study with instructive information. Despite the remarkable progress achieved in medical image segmentation, there is still an absence of 3D foundation segmentation model that can segment a wide range of anatomical categories with easy user interaction. In this paper, we propose a 3D foundation segmentation model, named SegVol, supporting universal and interactive volumetric medical image segmentation. By scaling up training data to 90k unlabeled Computed Tomography (CT) volumes and 6k labeled CT volumes, this foundation model supports the segmentation of over 200 anatomical categories using semantic and spatial prompts. Extensive experiments on 10 internal validation tasks and 18 external validation tasks verify that SegVol outperforms the state of the art by a large margin. Through its capacity to provide precise volumetric segmentation across various anatomical categories, SegVol has the potential to accelerate advancements in medical imaging diagnosis and facilitate treatment optimization.
Authors: Vasilis Siomos, Sergio Naval-Marimont, Jonathan Passerat-Palmbach, Giacomo Tarroni
Abstract: Federated Learning (FL) is a collaborative training paradigm that allows for privacy-preserving learning of cross-institutional models by eliminating the exchange of sensitive data and instead relying on the exchange of model parameters between the clients and a server. Despite individual studies on how client models are aggregated, and, more recently, on the benefits of ImageNet pre-training, there is a lack of understanding of the effect the architecture chosen for the federation has, and of how the aforementioned elements interconnect. To this end, we conduct the first joint ARchitecture-Initialization-Aggregation study and benchmark ARIAs across a range of medical image classification tasks. We find that, contrary to current practices, ARIA elements have to be chosen together to achieve the best possible performance. Our results also shed light on good choices for each element depending on the task, the effect of normalisation layers, and the utility of SSL pre-training, pointing to potential directions for designing FL-specific architectures and training pipelines.
Authors: Quoc Hung T. Le, Hieu H. Pham
Abstract: Musculoskeletal diseases and cognitive impairments in patients lead to difficulties in movement as well as negative effects on their psychological health. Clinical gait analysis, a vital tool for early diagnosis and treatment, traditionally relies on expensive optical motion capture systems. Recent advances in computer vision and deep learning have opened the door to more accessible and cost-effective alternatives. This paper introduces a novel spatio-temporal Transformer network to estimate critical gait parameters from RGB videos captured by a single-view camera. Empirical evaluations on a public dataset of cerebral palsy patients indicate that the proposed framework surpasses current state-of-the-art approaches and show significant improvements in predicting general gait parameters (including Walking Speed, Gait Deviation Index - GDI, and Knee Flexion Angle at Maximum Extension), while utilizing fewer parameters and alleviating the need for manual feature extraction.
Authors: Parul Gupta, Tuan Nguyen, Abhinav Dhall, Munawar Hayat, Trung Le, Thanh-Toan Do
Abstract: The task of Visual Relationship Recognition (VRR) aims to identify relationships between two interacting objects in an image and is particularly challenging due to the widely-spread and highly imbalanced distribution of
Authors: Peter Lorenz, Ricard Durall, Janis Keuper
Abstract: In recent years, diffusion models (DMs) have drawn significant attention for their success in approximating data distributions, yielding state-of-the-art generative results. Nevertheless, the versatility of these models extends beyond their generative capabilities to encompass various vision applications, such as image inpainting, segmentation, adversarial robustness, among others. This study is dedicated to the investigation of adversarial attacks through the lens of diffusion models. However, our objective does not involve enhancing the adversarial robustness of image classifiers. Instead, our focus lies in utilizing the diffusion model to detect and analyze the anomalies introduced by these attacks on images. To that end, we systematically examine the alignment of the distributions of adversarial examples when subjected to the process of transformation using diffusion models. The efficacy of this approach is assessed across CIFAR-10 and ImageNet datasets, including varying image sizes in the latter. The results demonstrate a notable capacity to discriminate effectively between benign and attacked images, providing compelling evidence that adversarial instances do not align with the learned manifold of the DMs.
Authors: Sina Borzooei, Leonardo Scabini, Gisele Miranda, Saba Daneshgar, Lukas Deblieck, Piet De Langhe, Odemir Bruno, Bernard De Baets, Ingmar Nopens, Elena Torfs
Abstract: Microbial communities play a key role in biological wastewater treatment processes. Activated sludge settling characteristics, for example, are affected by microbial community composition, varying by changes in operating conditions and influent characteristics of wastewater treatment plants (WWTPs). Timely assessment and prediction of changes in microbial composition leading to settling problems, such as filamentous bulking (FB), can prevent operational challenges, reductions in treatment efficiency, and adverse environmental impacts. This study presents an innovative computer vision-based approach to assess activated sludge-settling characteristics based on the morphological properties of flocs and filaments in microscopy images. Implementing the transfer learning of deep convolutional neural network (CNN) models, this approach aims to overcome the limitations of existing quantitative image analysis techniques. The offline microscopy image dataset was collected over two years, with weekly sampling at a full-scale industrial WWTP in Belgium. Multiple data augmentation techniques were employed to enhance the generalizability of the CNN models. Various CNN architectures, including Inception v3, ResNet18, ResNet152, ConvNeXt-nano, and ConvNeXt-S, were tested to evaluate their performance in predicting sludge settling characteristics. The sludge volume index was used as the final prediction variable, but the method can easily be adjusted to predict any other settling metric of choice. The results showed that the suggested CNN-based approach provides less labour-intensive, objective, and consistent assessments, while transfer learning notably minimises the training phase, resulting in a generalizable system that can be employed in real-time applications.
Authors: Mohsen Azarmi, Mahdi Rezaei, He Wang, Sebastien Glaser
Abstract: Accurate pedestrian intention prediction (PIP) by Autonomous Vehicles (AVs) is one of the current research challenges in this field. In this article, we introduce PIP-Net, a novel framework designed to predict pedestrian crossing intentions by AVs in real-world urban scenarios. We offer two variants of PIP-Net designed for different camera mounts and setups. Leveraging both kinematic data and spatial features from the driving scene, the proposed model employs a recurrent and temporal attention-based solution, outperforming state-of-the-art performance. To enhance the visual representation of road users and their proximity to the ego vehicle, we introduce a categorical depth feature map, combined with a local motion flow feature, providing rich insights into the scene dynamics. Additionally, we explore the impact of expanding the camera's field of view, from one to three cameras surrounding the ego vehicle, leading to enhancement in the model's contextual perception. Depending on the traffic scenario and road environment, the model excels in predicting pedestrian crossing intentions up to 4 seconds in advance which is a breakthrough in current research studies in pedestrian intention prediction. Finally, for the first time, we present the Urban-PIP dataset, a customised pedestrian intention prediction dataset, with multi-camera annotations in real-world automated driving scenarios.
Authors: Loddo Fabio, Dario Piga, Michelucci Umberto, El Ghazouali Safouane
Abstract: Satellites equipped with optical sensors capture high-resolution imagery, providing valuable insights into various environmental phenomena. In recent years, there has been a surge of research focused on addressing some challenges in remote sensing, ranging from water detection in diverse landscapes to the segmentation of mountainous and terrains. Ongoing investigations goals to enhance the precision and efficiency of satellite imagery analysis. Especially, there is a growing emphasis on developing methodologies for accurate water body detection, snow and clouds, important for environmental monitoring, resource management, and disaster response. Within this context, this paper focus on the cloud segmentation from remote sensing imagery. Accurate remote sensing data analysis can be challenging due to the presence of clouds in optical sensor-based applications. The quality of resulting products such as applications and research is directly impacted by cloud detection, which plays a key role in the remote sensing data processing pipeline. This paper examines seven cutting-edge semantic segmentation and detection algorithms applied to clouds identification, conducting a benchmark analysis to evaluate their architectural approaches and identify the most performing ones. To increase the model's adaptability, critical elements including the type of imagery and the amount of spectral bands used during training are analyzed. Additionally, this research tries to produce machine learning algorithms that can perform cloud segmentation using only a few spectral bands, including RGB and RGBN-IR combinations. The model's flexibility for a variety of applications and user scenarios is assessed by using imagery from Sentinel-2 and Landsat-8 as datasets. This benchmark can be reproduced using the material from this github link: https://github.com/toelt-llc/cloud_segmentation_comparative.
URLs: https://github.com/toelt-llc/cloud_segmentation_comparative.
Authors: Ashish Kumar, Laxmidhar Behera
Abstract: Autonomous aerial harvesting is a highly complex problem because it requires numerous interdisciplinary algorithms to be executed on mini low-powered computing devices. Object detection is one such algorithm that is compute-hungry. In this context, we make the following contributions: (i) Fast Fruit Detector (FFD), a resource-efficient, single-stage, and postprocessing-free object detector based on our novel latent object representation (LOR) module, query assignment, and prediction strategy. FFD achieves 100FPS@FP32 precision on the latest 10W NVIDIA Jetson-NX embedded device while co-existing with other time-critical sub-systems such as control, grasping, SLAM, a major achievement of this work. (ii) a method to generate vast amounts of training data without exhaustive manual labelling of fruit images since they consist of a large number of instances, which increases the labelling cost and time. (iii) an open-source fruit detection dataset having plenty of very small-sized instances that are difficult to detect. Our exhaustive evaluations on our and MinneApple dataset show that FFD, being only a single-scale detector, is more accurate than many representative detectors, e.g. FFD is better than single-scale Faster-RCNN by 10.7AP, multi-scale Faster-RCNN by 2.3AP, and better than latest single-scale YOLO-v8 by 8AP and multi-scale YOLO-v8 by 0.3 while being considerably faster.
Authors: Santiago Castro, Amir Ziai, Avneesh Saluja, Zhuoning Yuan, Rada Mihalcea
Abstract: Recent years have witnessed a significant increase in the performance of Vision and Language tasks. Foundational Vision-Language Models (VLMs), such as CLIP, have been leveraged in multiple settings and demonstrated remarkable performance across several tasks. Such models excel at object-centric recognition yet learn text representations that seem invariant to word order, failing to compose known concepts in novel ways. However, no evidence exists that any VLM, including large-scale single-stream models such as GPT-4V, identifies compositions successfully. In this paper, we introduce a framework to significantly improve the ability of existing models to encode compositional language, with over 10% absolute improvement on compositionality benchmarks, while maintaining or improving the performance on standard object-recognition and retrieval benchmarks. Our code and pre-trained models are publicly available at https://github.com/netflix/clove.
Authors: Jiazhao Zhang, Kunyu Wang, Rongtao Xu, Gengze Zhou, Yicong Hong, Xiaomeng Fang, Qi Wu, Zhizheng Zhang, Wang He
Abstract: Vision-and-Language Navigation (VLN) stands as a key research problem of Embodied AI, aiming at enabling agents to navigate in unseen environments following linguistic instructions. In this field, generalization is a long-standing challenge, either to out-of-distribution scenes or from Sim to Real. In this paper, we propose NaVid, a video-based large vision language model (VLM), to mitigate such a generalization gap. NaVid makes the first endeavour to showcase the capability of VLMs to achieve state-of-the-art level navigation performance without any maps, odometer and depth inputs. Following human instruction, NaVid only requires an on-the-fly video stream from a monocular RGB camera equipped on the robot to output the next-step action. Our formulation mimics how humans navigate and naturally gets rid of the problems introduced by odometer noises, and the Sim2Real gaps from map or depth inputs. Moreover, our video-based approach can effectively encode the historical observations of robots as spatio-temporal contexts for decision-making and instruction following. We train NaVid with 550k navigation samples collected from VLN-CE trajectories, including action-planning and instruction-reasoning samples, along with 665k large-scale web data. Extensive experiments show that NaVid achieves SOTA performance in simulation environments and the real world, demonstrating superior cross-dataset and Sim2Real transfer. We thus believe our proposed VLM approach plans the next step for not only the navigation agents but also this research field.
Authors: Chaokang Jiang, Guangming Wang, Jiuming Liu, Hesheng Wang, Zhuang Ma, Zhenqiang Liu, Zhujin Liang, Yi Shan, Dalong Du
Abstract: Learning 3D scene flow from LiDAR point clouds presents significant difficulties, including poor generalization from synthetic datasets to real scenes, scarcity of real-world 3D labels, and poor performance on real sparse LiDAR point clouds. We present a novel approach from the perspective of auto-labelling, aiming to generate a large number of 3D scene flow pseudo labels for real-world LiDAR point clouds. Specifically, we employ the assumption of rigid body motion to simulate potential object-level rigid movements in autonomous driving scenarios. By updating different motion attributes for multiple anchor boxes, the rigid motion decomposition is obtained for the whole scene. Furthermore, we developed a novel 3D scene flow data augmentation method for global and local motion. By perfectly synthesizing target point clouds based on augmented motion parameters, we easily obtain lots of 3D scene flow labels in point clouds highly consistent with real scenarios. On multiple real-world datasets including LiDAR KITTI, nuScenes, and Argoverse, our method outperforms all previous supervised and unsupervised methods without requiring manual labelling. Impressively, our method achieves a tenfold reduction in EPE3D metric on the LiDAR KITTI dataset, reducing it from $0.190m$ to a mere $0.008m$ error.
Authors: Weilin Wan, Weizhong Zhang, Cheng Jin
Abstract: Out-of-distribution detection is a crucial technique for deploying machine learning models in the real world to handle the unseen scenarios. In this paper, we propose a simple but effective Neural Activation Prior (NAP) for out-of-distribution detection (OOD). Our neural activation prior is based on a key observation that, for a channel before the global pooling layer of a fully trained neural network, the probability of a few of its neurons being activated with a larger response by an in-distribution (ID) sample is significantly higher than that by an OOD sample. An intuitive explanation is each channel in a model fully trained on ID dataset would play a role in detecting a certain pattern in the samples within the ID dataset, and a few neurons can be activated with a large response when the pattern is detected in an input sample. Thus, a new scoring function based on this prior is proposed to highlight the role of these strongly activated neurons in OOD detection. This approach is plug-and-play and does not lead to any performance degradation on in-distribution data classification and requires no extra training or statistics from training or external datasets. Notice that previous methods primarily rely on post-global-pooling features of the neural networks, while the within-channel distribution information we leverage would be discarded by the global pooling operator. Consequently, our method is orthogonal to existing approaches and can be effectively combined with them in various applications. Experimental results show that our method achieves the state-of-the-art performance on CIFAR-10, CIFAR-100 and ImageNet datasets, which demonstrates the power of the proposed prior.
Authors: Zhengqing Zang, Chenyu Lin, Chenwei Tang, Tao Wang, Jiancheng Lv
Abstract: Existing object detection models are mainly trained on large-scale labeled datasets. However, annotating data for novel aerial object classes is expensive since it is time-consuming and may require expert knowledge. Thus, it is desirable to study label-efficient object detection methods on aerial images. In this work, we propose a zero-shot method for aerial object detection named visual Description Regularization, or DescReg. Concretely, we identify the weak semantic-visual correlation of the aerial objects and aim to address the challenge with prior descriptions of their visual appearance. Instead of directly encoding the descriptions into class embedding space which suffers from the representation gap problem, we propose to infuse the prior inter-class visual similarity conveyed in the descriptions into the embedding learning. The infusion process is accomplished with a newly designed similarity-aware triplet loss which incorporates structured regularization on the representation space. We conduct extensive experiments with three challenging aerial object detection datasets, including DIOR, xView, and DOTA. The results demonstrate that DescReg significantly outperforms the state-of-the-art ZSD methods with complex projection designs and generative frameworks, e.g., DescReg outperforms best reported ZSD method on DIOR by 4.5 mAP on unseen classes and 8.1 in HM. We further show the generalizability of DescReg by integrating it into generative ZSD methods as well as varying the detection architecture.
Authors: Hongjun Wang, Jiyuan Chen, Yinqiang Zheng, Tieyong Zeng
Abstract: Deep learning has led to a dramatic leap on Single Image Super-Resolution (SISR) performances in recent years. %Despite the substantial advancement% While most existing work assumes a simple and fixed degradation model (e.g., bicubic downsampling), the research of Blind SR seeks to improve model generalization ability with unknown degradation. Recently, Kong et al pioneer the investigation of a more suitable training strategy for Blind SR using Dropout. Although such method indeed brings substantial generalization improvements via mitigating overfitting, we argue that Dropout simultaneously introduces undesirable side-effect that compromises model's capacity to faithfully reconstruct fine details. We show both the theoretical and experimental analyses in our paper, and furthermore, we present another easy yet effective training strategy that enhances the generalization ability of the model by simply modulating its first and second-order features statistics. Experimental results have shown that our method could serve as a model-agnostic regularization and outperforms Dropout on seven benchmark datasets including both synthetic and real-world scenarios.
Authors: Niccol\`o Cavagnero, Gabriele Rosi, Claudia Cuttano, Francesca Pistilli, Marco Ciccone, Giuseppe Averta, Fabio Cermelli
Abstract: Recent transformer-based architectures have shown impressive results in the field of image segmentation. Thanks to their flexibility, they obtain outstanding performance in multiple segmentation tasks, such as semantic and panoptic, under a single unified framework. To achieve such impressive performance, these architectures employ intensive operations and require substantial computational resources, which are often not available, especially on edge devices. To fill this gap, we propose Prototype-based Efficient MaskFormer (PEM), an efficient transformer-based architecture that can operate in multiple segmentation tasks. PEM proposes a novel prototype-based cross-attention which leverages the redundancy of visual features to restrict the computation and improve the efficiency without harming the performance. In addition, PEM introduces an efficient multi-scale feature pyramid network, capable of extracting features that have high semantic content in an efficient way, thanks to the combination of deformable convolutions and context-based self-modulation. We benchmark the proposed PEM architecture on two tasks, semantic and panoptic segmentation, evaluated on two different datasets, Cityscapes and ADE20K. PEM demonstrates outstanding performance on every task and dataset, outperforming task-specific architectures while being comparable and even better than computationally-expensive baselines.
Authors: Jinghuai Zhang, Hongbin Liu, Jinyuan Jia, Neil Zhenqiang Gong
Abstract: Contrastive learning (CL) pre-trains general-purpose encoders using an unlabeled pre-training dataset, which consists of images or image-text pairs. CL is vulnerable to data poisoning based backdoor attacks (DPBAs), in which an attacker injects poisoned inputs into the pre-training dataset so the encoder is backdoored. However, existing DPBAs achieve limited effectiveness. In this work, we take the first step to analyze the limitations of existing backdoor attacks and propose new DPBAs called CorruptEncoder to CL. CorruptEncoder introduces a new attack strategy to create poisoned inputs and uses a theory-guided method to maximize attack effectiveness. Our experiments show that CorruptEncoder substantially outperforms existing DPBAs. In particular, CorruptEncoder is the first DPBA that achieves more than 90% attack success rates with only a few (3) reference images and a small poisoning ratio 0.5%. Moreover, we also propose a defense, called localized cropping, to defend against DPBAs. Our results show that our defense can reduce the effectiveness of DPBAs, but it sacrifices the utility of the encoder, highlighting the need for new defenses.
Authors: Sam Bond-Taylor, Chris G. Willcocks
Abstract: This paper introduces $\infty$-Diff, a generative diffusion model defined in an infinite-dimensional Hilbert space, which can model infinite resolution data. By training on randomly sampled subsets of coordinates and denoising content only at those locations, we learn a continuous function for arbitrary resolution sampling. Unlike prior neural field-based infinite-dimensional models, which use point-wise functions requiring latent compression, our method employs non-local integral operators to map between Hilbert spaces, allowing spatial context aggregation. This is achieved with an efficient multi-scale function-space architecture that operates directly on raw sparse coordinates, coupled with a mollified diffusion process that smooths out irregularities. Through experiments on high-resolution datasets, we found that even at an $8\times$ subsampling rate, our model retains high-quality diffusion. This leads to significant run-time and memory savings, delivers samples with lower FID scores, and scales beyond the training resolution while retaining detail.
Authors: Sara Major, Aleksandar Toma\v{s}evi\'c
Abstract: Online media has revolutionized the way political information is disseminated and consumed on a global scale, and this shift has compelled political figures to adopt new strategies of capturing and retaining voter attention. These strategies often rely on emotional persuasion and appeal, and as visual content becomes increasingly prevalent in virtual space, much of political communication too has come to be marked by evocative video content and imagery. The present paper offers a novel approach to analyzing material of this kind. We apply a deep-learning-based computer-vision algorithm to a sample of 220 YouTube videos depicting political leaders from 15 different countries, which is based on an existing trained convolutional neural network architecture provided by the Python library fer. The algorithm returns emotion scores representing the relative presence of 6 emotional states (anger, disgust, fear, happiness, sadness, and surprise) and a neutral expression for each frame of the processed YouTube video. We observe statistically significant differences in the average score of expressed negative emotions between groups of leaders with varying degrees of populist rhetoric as defined by the Global Party Survey (GPS), indicating that populist leaders tend to express negative emotions to a greater extent during their public performance than their non-populist counterparts. Overall, our contribution provides insight into the characteristics of visual self-representation among political leaders, as well as an open-source workflow for further computational studies of their non-verbal communication.
Authors: Rong Ma, Eric D. Sun, David Donoho, James Zou
Abstract: Single-cell data integration can provide a comprehensive molecular view of cells, and many algorithms have been developed to remove unwanted technical or biological variations and integrate heterogeneous single-cell datasets. Despite their wide usage, existing methods suffer from several fundamental limitations. In particular, we lack a rigorous statistical test for whether two high-dimensional single-cell datasets are alignable (and therefore should even be aligned). Moreover, popular methods can substantially distort the data during alignment, making the aligned data and downstream analysis difficult to interpret. To overcome these limitations, we present a spectral manifold alignment and inference (SMAI) framework, which enables principled and interpretable alignability testing and structure-preserving integration of single-cell data with the same type of features. SMAI provides a statistical test to robustly assess the alignability between datasets to avoid misleading inference, and is justified by high-dimensional statistical theory. On a diverse range of real and simulated benchmark datasets, it outperforms commonly used alignment methods. Moreover, we show that SMAI improves various downstream analyses such as identification of differentially expressed genes and imputation of single-cell spatial transcriptomics, providing further biological insights. SMAI's interpretability also enables quantification and a deeper understanding of the sources of technical confounders in single-cell data.
Authors: Haomin Zhuang, Mingxian Yu, Hao Wang, Yang Hua, Jian Li, Xu Yuan
Abstract: Federated learning (FL) has been widely deployed to enable machine learning training on sensitive data across distributed devices. However, the decentralized learning paradigm and heterogeneity of FL further extend the attack surface for backdoor attacks. Existing FL attack and defense methodologies typically focus on the whole model. None of them recognizes the existence of backdoor-critical (BC) layers-a small subset of layers that dominate the model vulnerabilities. Attacking the BC layers achieves equivalent effects as attacking the whole model but at a far smaller chance of being detected by state-of-the-art (SOTA) defenses. This paper proposes a general in-situ approach that identifies and verifies BC layers from the perspective of attackers. Based on the identified BC layers, we carefully craft a new backdoor attack methodology that adaptively seeks a fundamental balance between attacking effects and stealthiness under various defense strategies. Extensive experiments show that our BC layer-aware backdoor attacks can successfully backdoor FL under seven SOTA defenses with only 10% malicious clients and outperform the latest backdoor attack methods.
Authors: A. Abdullah, T. Barua, R. Tibbetts, Z. Chen, M. J. Islam, I. Rekleitis
Abstract: In this paper, we present CaveSeg - the first visual learning pipeline for semantic segmentation and scene parsing for AUV navigation inside underwater caves. We address the problem of scarce annotated training data by preparing a comprehensive dataset for semantic segmentation of underwater cave scenes. It contains pixel annotations for important navigation markers (e.g. caveline, arrows), obstacles (e.g. ground plain and overhead layers), scuba divers, and open areas for servoing. Through comprehensive benchmark analyses on cave systems in USA, Mexico, and Spain locations, we demonstrate that robust deep visual models can be developed based on CaveSeg for fast semantic scene parsing of underwater cave environments. In particular, we formulate a novel transformer-based model that is computationally light and offers near real-time execution in addition to achieving state-of-the-art performance. Finally, we explore the design choices and implications of semantic segmentation for visual servoing by AUVs inside underwater caves. The proposed model and benchmark dataset open up promising opportunities for future research in autonomous underwater cave exploration and mapping.
Authors: Qiaojun Yu, Junbo Wang, Wenhai Liu, Ce Hao, Liu Liu, Lin Shao, Weiming Wang, Cewu Lu
Abstract: Articulated objects like cabinets and doors are widespread in daily life. However, directly manipulating 3D articulated objects is challenging because they have diverse geometrical shapes, semantic categories, and kinetic constraints. Prior works mostly focused on recognizing and manipulating articulated objects with specific joint types. They can either estimate the joint parameters or distinguish suitable grasp poses to facilitate trajectory planning. Although these approaches have succeeded in certain types of articulated objects, they lack generalizability to unseen objects, which significantly impedes their application in broader scenarios. In this paper, we propose a novel framework of Generalizable Articulation Modeling and Manipulating for Articulated Objects (GAMMA), which learns both articulation modeling and grasp pose affordance from diverse articulated objects with different categories. In addition, GAMMA adopts adaptive manipulation to iteratively reduce the modeling errors and enhance manipulation performance. We train GAMMA with the PartNet-Mobility dataset and evaluate with comprehensive experiments in SAPIEN simulation and real-world Franka robot. Results show that GAMMA significantly outperforms SOTA articulation modeling and manipulation algorithms in unseen and cross-category articulated objects. We will open-source all codes and datasets in both simulation and real robots for reproduction in the final version. Images and videos are published on the project website at: http://sites.google.com/view/gamma-articulation
Authors: Shih-Yang Liu, Chien-Yi Wang, Hongxu Yin, Pavlo Molchanov, Yu-Chiang Frank Wang, Kwang-Ting Cheng, Min-Hung Chen
Abstract: Among the widely used parameter-efficient finetuning (PEFT) methods, LoRA and its variants have gained considerable popularity because of avoiding additional inference costs. However, there still often exists an accuracy gap between these methods and full fine-tuning (FT). In this work, we first introduce a novel weight decomposition analysis to investigate the inherent differences between FT and LoRA. Aiming to resemble the learning capacity of FT from the findings, we propose Weight-Decomposed LowRank Adaptation (DoRA). DoRA decomposes the pre-trained weight into two components, magnitude and direction, for fine-tuning, specifically employing LoRA for directional updates to efficiently minimize the number of trainable parameters. By employing DoRA, we enhance both the learning capacity and training stability of LoRA while avoiding any additional inference overhead. DoRA consistently outperforms LoRA on fine-tuning LLaMA, LLaVA, and VL-BART on various downstream tasks, such as commonsense reasoning, visual instruction tuning, and image/video-text understanding.
Authors: Sarina Thomas, Cristiana Tiago, B{\o}rge Solli Andreassen, Svein Arne Aase, Jurica \v{S}prem, Erik Steen, Anne Solberg, Guy Ben-Yosef
Abstract: To facilitate diagnosis on cardiac ultrasound (US), clinical practice has established several standard views of the heart, which serve as reference points for diagnostic measurements and define viewports from which images are acquired. Automatic view recognition involves grouping those images into classes of standard views. Although deep learning techniques have been successful in achieving this, they still struggle with fully verifying the suitability of an image for specific measurements due to factors like the correct location, pose, and potential occlusions of cardiac structures. Our approach goes beyond view classification and incorporates a 3D mesh reconstruction of the heart that enables several more downstream tasks, like segmentation and pose estimation. In this work, we explore learning 3D heart meshes via graph convolutions, using similar techniques to learn 3D meshes in natural images, such as human pose estimation. As the availability of fully annotated 3D images is limited, we generate synthetic US images from 3D meshes by training an adversarial denoising diffusion model. Experiments were conducted on synthetic and clinical cases for view recognition and structure detection. The approach yielded good performance on synthetic images and, despite being exclusively trained on synthetic data, it already showed potential when applied to clinical images. With this proof-of-concept, we aim to demonstrate the benefits of graphs to improve cardiac view recognition that can ultimately lead to better efficiency in cardiac diagnosis.
Authors: Kate Sanders, Nathaniel Weir, Benjamin Van Durme
Abstract: It is challenging to perform question-answering over complex, multimodal content such as television clips. This is in part because current video-language models rely on single-modality reasoning, have lowered performance on long inputs, and lack interpetability. We propose TV-TREES, the first multimodal entailment tree generator. TV-TREES serves as an approach to video understanding that promotes interpretable joint-modality reasoning by producing trees of entailment relationships between simple premises directly entailed by the videos and higher-level conclusions. We then introduce the task of multimodal entailment tree generation to evaluate the reasoning quality of such methods. Our method's experimental results on the challenging TVQA dataset demonstrate intepretable, state-of-the-art zero-shot performance on full video clips, illustrating a best of both worlds contrast to black-box methods.