David Doermann

Mahesh Bhosale, Naresh Kumar Devulapally, Abdul Wasi et al.

Diffusion models have emerged as the backbone of modern generative AI, powering advances in vision, language, audio and other modalities. Despite their success, they suffer from hallucinations, implausible samples that lie outside the support of true data distribution, which degrade reliability and trust. In this work, we first empirically confirm previously proposed hypothesis that score smoothness causes hallucinations in Image Generation diffusion models and provide a density-based perspective. We further formalize this notion by linking the hallucinations probability mass to lipschitz constant of the learned score function. Motivated by this, we introduce a Variance-Guided Score Modulation (VSM) strategy that controls the score Jacobian, in turn reducing score smoothness and better approximating the ground truth score that decreases hallucinations. Empirical results on synthetic and real-world datasets demonstrate that our approach reduces hallucinations (up to ~25%) while maintaining high fidelity and diversity, providing a principled step toward more reliable diffusion-based image generation. We also propose two benchmark datasets with extreme semantic variation for systematic hallucination evaluation. Code and Datasets are publicly available at https://github.com/bhosalems/VSM.

Shaohui Lin, Wenxuan Huang, Jiao Xie et al.

Filter pruning simultaneously accelerates the computation and reduces the memory overhead of CNNs, which can be effectively applied to edge devices and cloud services. In this paper, we propose a novel Knowledge-driven Differential Filter Sampler~(KDFS) with Masked Filter Modeling~(MFM) framework for filter pruning, which globally prunes the redundant filters based on the prior knowledge of a pre-trained model in a differential and non-alternative optimization. Specifically, we design a differential sampler with learnable sampling parameters to build a binary mask vector for each layer, determining whether the corresponding filters are redundant. To learn the mask, we introduce masked filter modeling to construct PCA-like knowledge by aligning the intermediate features from the pre-trained teacher model and the outputs of the student decoder taking sampling features as the input. The mask and sampler are directly optimized by the Gumbel-Softmax Straight-Through Gradient Estimator in an end-to-end manner in combination with global pruning constraint, MFM reconstruction error, and dark knowledge. Extensive experiments demonstrate the proposed KDFS's effectiveness in compressing the base models on various datasets. For instance, the pruned ResNet-50 on ImageNet achieves $55.36\%$ computation reduction, and $42.86\%$ parameter reduction, while only dropping $0.35\%$ Top-1 accuracy, significantly outperforming the state-of-the-art methods. The code is available at \url{https://github.com/Osilly/KDFS}.

Yuanhao Zhai, Kevin Lin, Linjie Li et al. · microsoft-research

Significant advances have been made in human-centric video generation, yet the joint video-depth generation problem remains underexplored. Most existing monocular depth estimation methods may not generalize well to synthesized images or videos, and multi-view-based methods have difficulty controlling the human appearance and motion. In this work, we present IDOL (unIfied Dual-mOdal Latent diffusion) for high-quality human-centric joint video-depth generation. Our IDOL consists of two novel designs. First, to enable dual-modal generation and maximize the information exchange between video and depth generation, we propose a unified dual-modal U-Net, a parameter-sharing framework for joint video and depth denoising, wherein a modality label guides the denoising target, and cross-modal attention enables the mutual information flow. Second, to ensure a precise video-depth spatial alignment, we propose a motion consistency loss that enforces consistency between the video and depth feature motion fields, leading to harmonized outputs. Additionally, a cross-attention map consistency loss is applied to align the cross-attention map of the video denoising with that of the depth denoising, further facilitating spatial alignment. Extensive experiments on the TikTok and NTU120 datasets show our superior performance, significantly surpassing existing methods in terms of video FVD and depth accuracy.

Xuan Gong, Abhishek Sharma, Srikrishna Karanam et al.

Federated Learning (FL) is a machine learning paradigm where local nodes collaboratively train a central model while the training data remains decentralized. Existing FL methods typically share model parameters or employ co-distillation to address the issue of unbalanced data distribution. However, they suffer from communication bottlenecks. More importantly, they risk privacy leakage. In this work, we develop a privacy preserving and communication efficient method in a FL framework with one-shot offline knowledge distillation using unlabeled, cross-domain public data. We propose a quantized and noisy ensemble of local predictions from completely trained local models for stronger privacy guarantees without sacrificing accuracy. Based on extensive experiments on image classification and text classification tasks, we show that our privacy-preserving method outperforms baseline FL algorithms with superior performance in both accuracy and communication efficiency.

Xuan Gong, Liangchen Song, Rishi Vedula et al.

Federated Learning (FL) is a machine learning paradigm where many local nodes collaboratively train a central model while keeping the training data decentralized. This is particularly relevant for clinical applications since patient data are usually not allowed to be transferred out of medical facilities, leading to the need for FL. Existing FL methods typically share model parameters or employ co-distillation to address the issue of unbalanced data distribution. However, they also require numerous rounds of synchronized communication and, more importantly, suffer from a privacy leakage risk. We propose a privacy-preserving FL framework leveraging unlabeled public data for one-way offline knowledge distillation in this work. The central model is learned from local knowledge via ensemble attention distillation. Our technique uses decentralized and heterogeneous local data like existing FL approaches, but more importantly, it significantly reduces the risk of privacy leakage. We demonstrate that our method achieves very competitive performance with more robust privacy preservation based on extensive experiments on image classification, segmentation, and reconstruction tasks.

Yuanhao Zhai, Mingzhen Huang, Tianyu Luan et al.

Language-guided human motion synthesis has been a challenging task due to the inherent complexity and diversity of human behaviors. Previous methods face limitations in generalization to novel actions, often resulting in unrealistic or incoherent motion sequences. In this paper, we propose ATOM (ATomic mOtion Modeling) to mitigate this problem, by decomposing actions into atomic actions, and employing a curriculum learning strategy to learn atomic action composition. First, we disentangle complex human motions into a set of atomic actions during learning, and then assemble novel actions using the learned atomic actions, which offers better adaptability to new actions. Moreover, we introduce a curriculum learning training strategy that leverages masked motion modeling with a gradual increase in the mask ratio, and thus facilitates atomic action assembly. This approach mitigates the overfitting problem commonly encountered in previous methods while enforcing the model to learn better motion representations. We demonstrate the effectiveness of ATOM through extensive experiments, including text-to-motion and action-to-motion synthesis tasks. We further illustrate its superiority in synthesizing plausible and coherent text-guided human motion sequences.

Liangchen Song, Xuan Gong, Benjamin Planche et al.

Knowing the 3D motions in a dynamic scene is essential to many vision applications. Recent progress is mainly focused on estimating the activity of some specific elements like humans. In this paper, we leverage a neural motion field for estimating the motion of all points in a multiview setting. Modeling the motion from a dynamic scene with multiview data is challenging due to the ambiguities in points of similar color and points with time-varying color. We propose to regularize the estimated motion to be predictable. If the motion from previous frames is known, then the motion in the near future should be predictable. Therefore, we introduce a predictability regularization by first conditioning the estimated motion on latent embeddings, then by adopting a predictor network to enforce predictability on the embeddings. The proposed framework PREF (Predictability REgularized Fields) achieves on par or better results than state-of-the-art neural motion field-based dynamic scene representation methods, while requiring no prior knowledge of the scene.

Xuan Gong, Liangchen Song, Meng Zheng et al.

To date, little attention has been given to multi-view 3D human mesh estimation, despite real-life applicability (e.g., motion capture, sport analysis) and robustness to single-view ambiguities. Existing solutions typically suffer from poor generalization performance to new settings, largely due to the limited diversity of image-mesh pairs in multi-view training data. To address this shortcoming, people have explored the use of synthetic images. But besides the usual impact of visual gap between rendered and target data, synthetic-data-driven multi-view estimators also suffer from overfitting to the camera viewpoint distribution sampled during training which usually differs from real-world distributions. Tackling both challenges, we propose a novel simulation-based training pipeline for multi-view human mesh recovery, which (a) relies on intermediate 2D representations which are more robust to synthetic-to-real domain gap; (b) leverages learnable calibration and triangulation to adapt to more diversified camera setups; and (c) progressively aggregates multi-view information in a canonical 3D space to remove ambiguities in 2D representations. Through extensive benchmarking, we demonstrate the superiority of the proposed solution especially for unseen in-the-wild scenarios.

Xuan Gong, Meng Zheng, Benjamin Planche et al.

Fully supervised human mesh recovery methods are data-hungry and have poor generalizability due to the limited availability and diversity of 3D-annotated benchmark datasets. Recent progress in self-supervised human mesh recovery has been made using synthetic-data-driven training paradigms where the model is trained from synthetic paired 2D representation (e.g., 2D keypoints and segmentation masks) and 3D mesh. However, on synthetic dense correspondence maps (i.e., IUV) few have been explored since the domain gap between synthetic training data and real testing data is hard to address for 2D dense representation. To alleviate this domain gap on IUV, we propose cross-representation alignment utilizing the complementary information from the robust but sparse representation (2D keypoints). Specifically, the alignment errors between initial mesh estimation and both 2D representations are forwarded into regressor and dynamically corrected in the following mesh regression. This adaptive cross-representation alignment explicitly learns from the deviations and captures complementary information: robustness from sparse representation and richness from dense representation. We conduct extensive experiments on multiple standard benchmark datasets and demonstrate competitive results, helping take a step towards reducing the annotation effort needed to produce state-of-the-art models in human mesh estimation.

Saleem Ahmed, Pengyu Yan, David Doermann et al.

We introduce a novel bottom-up approach for the extraction of chart data. Our model utilizes images of charts as inputs and learns to detect keypoints (KP), which are used to reconstruct the components within the plot area. Our novelty lies in detecting a fusion of continuous and discrete KP as predicted heatmaps. A combination of sparse and dense per-pixel objectives coupled with a uni-modal self-attention-based feature-fusion layer is applied to learn KP embeddings. Further leveraging deep metric learning for unsupervised clustering, allows us to segment the chart plot area into various objects. By further matching the chart components to the legend, we are able to obtain the data series names. A post-processing threshold is applied to the KP embeddings to refine the object reconstructions and improve accuracy. Our extensive experiments include an evaluation of different modules for KP estimation and the combination of deep layer aggregation and corner pooling approaches. The results of our experiments provide extensive evaluation for the task of real-world chart data extraction.

Mahesh Bhosale, Abdul Wasi, Vishvesh Trivedi et al.

Grounded multi-video question answering over real-world news events requires systems to surface query-relevant evidence across heterogeneous video archives while attributing every claim to its supporting source. We introduce CRAFT (Critic-Refined Adaptive Key-Frame Targeting), a query-conditioned pipeline that combines dynamic keyframe selection, per-video ASR with multilingual fallback, and a hybrid critic loop to iteratively verify and repair claims before consolidation. The pipeline integrates UNLI temporal entailment, DeBERTa-v3 cross-claim screening, and a Llama-3.2-3B adjudicator, with a final citation-merging stage that emits each fact once with all supporting source identifiers. On MAGMaR 2026, CRAFT achieves the best overall average (0.739), reference recall (0.810), and citation F1 (0.635). We further evaluate on a MAGMaR-style conversion of WikiVideo with 52 non-overlapping event queries, where CRAFT also performs strongly (0.823 Avg), showing that its claim-centric evidence aggregation generalizes beyond MAGMaR. Ablations show that atomic claims, ASR, and the critic loop drive the main gains over the vanilla query-conditioned baseline. Code and implementation details are publicly available at https://github.com/bhosalems/CRAFT.

Runqi Wang, Linlin Yang, Baochang Zhang et al.

Research on the generalization ability of deep neural networks (DNNs) has recently attracted a great deal of attention. However, due to their complex architectures and large numbers of parameters, measuring the generalization ability of specific DNN models remains an open challenge. In this paper, we propose to use multiple factors to measure and rank the relative generalization of DNNs based on a new concept of confidence dimension (CD). Furthermore, we provide a feasible framework in our CD to theoretically calculate the upper bound of generalization based on the conventional Vapnik-Chervonenk dimension (VC-dimension) and Hoeffding's inequality. Experimental results on image classification and object detection demonstrate that our CD can reflect the relative generalization ability for different DNNs. In addition to full-precision DNNs, we also analyze the generalization ability of binary neural networks (BNNs), whose generalization ability remains an unsolved problem. Our CD yields a consistent and reliable measure and ranking for both full-precision DNNs and BNNs on all the tasks.

Mahesh Bhosale, Abdul Wasi, Shantam Srivastava et al.

While powerful in image-conditioned generation, multimodal large language models (MLLMs) can display uneven performance across demographic groups, highlighting fairness risks. In safety-critical clinical settings, such disparities risk producing unequal diagnostic narratives and eroding trust in AI-assisted decision-making. While fairness has been studied extensively in vision-only and language-only models, its impact on MLLMs remains largely underexplored. To address these biases, we introduce FairLLaVA, a parameter-efficient fine-tuning method that mitigates group disparities in visual instruction tuning without compromising overall performance. By minimizing the mutual information between target attributes, FairLLaVA regularizes the model's representations to be demographic-invariant. The method can be incorporated as a lightweight plug-in, maintaining efficiency with low-rank adapter fine-tuning, and provides an architecture-agnostic approach to fair visual instruction following. Extensive experiments on large-scale chest radiology report generation and dermoscopy visual question answering benchmarks show that FairLLaVA consistently reduces inter-group disparities while improving both equity-scaled clinical performance and natural language generation quality across diverse medical imaging modalities. Code can be accessed at https://github.com/bhosalems/FairLLaVA.

Yunjie Tian, Qixiang Ye, David Doermann

Enhancing the network architecture of the YOLO framework has been crucial for a long time, but has focused on CNN-based improvements despite the proven superiority of attention mechanisms in modeling capabilities. This is because attention-based models cannot match the speed of CNN-based models. This paper proposes an attention-centric YOLO framework, namely YOLOv12, that matches the speed of previous CNN-based ones while harnessing the performance benefits of attention mechanisms. YOLOv12 surpasses all popular real-time object detectors in accuracy with competitive speed. For example, YOLOv12-N achieves 40.6% mAP with an inference latency of 1.64 ms on a T4 GPU, outperforming advanced YOLOv10-N / YOLOv11-N by 2.1%/1.2% mAP with a comparable speed. This advantage extends to other model scales. YOLOv12 also surpasses end-to-end real-time detectors that improve DETR, such as RT-DETR / RT-DETRv2: YOLOv12-S beats RT-DETR-R18 / RT-DETRv2-R18 while running 42% faster, using only 36% of the computation and 45% of the parameters. More comparisons are shown in Figure 1.

Pengyu Yan, Akhil Gorugantu, Mahesh Bhosale et al.

Multi-video event understanding demands models that can locate and attribute query-relevant evidence scattered across long, heterogeneous video corpora. Existing large vision-language models (LVLMs) often underperform in this regime because they quickly exhaust their context budget and struggle to precisely localize evidentially important segments, frequently missing dense informational cues such as broadcast graphics, subtitles, and scoreboards. We introduce TRACE, an evidence grounding-guided framework that follows a ground-before-reasoning strategy for multi-video event reasoning. Our approach first builds a structured, text-searchable timeline for each video using OCR and object detection. A text-only LLM then conducts query-aware evidence localization, selecting relevant moments prior to any downstream visual reasoning. The retrieved frames and their grounding summaries are subsequently used to steer LVLM-based claim generation and cross-video citation consolidation. Experiments on MAGMaR 2026 and WikiVideo demonstrate that structured grounding markedly boosts factual completeness and attribution fidelity. On the MAGMaR validation split, TRACE raises macro-average MiRAGE F1 from 0.705 to 0.811 compared to an unguided Qwen3-VL-30B baseline, with especially strong improvements in citation recall from 0.440 to 0.628. The method also attains state-of-the-art results on the official MAGMaR 2026 leaderboard.

Tianyu Luan, Xuelu Feng, Zixin Zhu et al.

Textured high-fidelity 3D models are crucial for games, AR/VR, and film, but human-aligned evaluation methods still fall behind despite recent advances in 3D reconstruction and generation. Existing metrics, such as Chamfer Distance, often fail to align with how humans evaluate the fidelity of 3D shapes. Recent learning-based metrics attempt to improve this by relying on rendered images and 2D image quality metrics. However, these approaches face limitations due to incomplete structural coverage and sensitivity to viewpoint choices. Moreover, most methods are trained on synthetic distortions, which differ significantly from real-world distortions, resulting in a domain gap. To address these challenges, we propose a new fidelity evaluation method that is based directly on 3D meshes with texture, without relying on rendering. Our method, named Textured Geometry Evaluation TGE, jointly uses the geometry and color information to calculate the fidelity of the input textured mesh with comparison to a reference colored shape. To train and evaluate our metric, we design a human-annotated dataset with real-world distortions. Experiments show that TGE outperforms rendering-based and geometry-only methods on real-world distortion dataset.

Mahesh Bhosale, Abhishek Kumar, David Doermann

We propose a neural network architecture that learns body part appearances for soccer player re-identification. Our model consists of a two-stream network (one stream for appearance map extraction and the other for body part map extraction) and a bilinear-pooling layer that generates and spatially pools the body part map. Each local feature of the body part map is obtained by a bilinear mapping of the corresponding local appearance and body part descriptors. Our novel representation yields a robust image-matching feature map, which results from combining the local similarities of the relevant body parts with the weighted appearance similarity. Our model does not require any part annotation on the SoccerNet-V3 re-identification dataset to train the network. Instead, we use a sub-network of an existing pose estimation network (OpenPose) to initialize the part substream and then train the entire network to minimize the triplet loss. The appearance stream is pre-trained on the ImageNet dataset, and the part stream is trained from scratch for the SoccerNet-V3 dataset. We demonstrate the validity of our model by showing that it outperforms state-of-the-art models such as OsNet and InceptionNet.

Tongfei Chen, Jingying Yang, Linlin Yang et al.

Deep learning architectures are fundamentally inspired by neuroscience, particularly the structure of the brain's sensory pathways, and have achieved remarkable success in learning informative data representations. Although these architectures mimic the communication mechanisms of biological neurons, their strategies for information encoding and transmission are fundamentally distinct. Biological systems depend on dynamic fluctuations in membrane potential; by contrast, conventional deep networks optimize weights and biases by adjusting the strengths of inter-neural connections, lacking a systematic mechanism to jointly characterize the interplay among signal intensity, coupling structure, and state evolution. To tackle this limitation, we propose the Kirchhoff-Inspired Neural Network (KINN), a state-variable-based network architecture constructed based on Kirchhoff's current law. KINN derives numerically stable state updates from fundamental ordinary differential equations, enabling the explicit decoupling and encoding of higher-order evolutionary components within a single layer while preserving physical consistency, interpretability, and end-to-end trainability. Extensive experiments on partial differential equation (PDE) solving and ImageNet image classification validate that KINN outperforms state-of-the-art existing methods.

Shantam Srivastava, Mahesh Bhosale, David Doermann et al.

Interpreting chest X-rays is inherently challenging due to the overlap between anatomical structures and the subtle presentation of many clinically significant pathologies, making accurate diagnosis time-consuming even for experienced radiologists. Recent radiology-focused foundation models, such as LLaVA-Rad and Maira-2, have positioned multi-modal large language models (MLLMs) at the forefront of automated radiology report generation (RRG). However, despite these advances, current foundation models generate reports in a single forward pass. This decoding strategy diminishes attention to visual tokens and increases reliance on language priors as generation proceeds, which in turn introduces spurious pathology co-occurrences in the generated reports. To mitigate these limitations, we propose Category-Wise Contrastive Decoding (CWCD), a novel and modular framework designed to enhance structured radiology report generation (SRRG). Our approach introduces category-specific parameterization and generates category-wise reports by contrasting normal X-rays with masked X-rays using category-specific visual prompts. Experimental results demonstrate that CWCD consistently outperforms baseline methods across both clinical efficacy and natural language generation metrics. An ablation study further elucidates the contribution of each architectural component to overall performance.

Chau Pham, Quan Dao, Mahesh Bhosale et al.

Recent advances in diffusion models have revolutionized text-guided image editing, yet existing editing methods face critical challenges in hyperparameter identification. To get the reasonable editing performance, these methods often require the user to brute-force tune multiple interdependent hyperparameters, such as inversion timesteps and attention modification. This process incurs high computational costs due to the huge hyperparameter search space. We consider searching optimal editing's hyperparameters as a sequential decision-making task within the diffusion denoising process. Specifically, we propose a reinforcement learning framework, which establishes a Markov Decision Process that dynamically adjusts hyperparameters across denoising steps, integrating editing objectives into a reward function. The method achieves time efficiency through proximal policy optimization while maintaining optimal hyperparameter configurations. Experiments demonstrate significant reduction in search time and computational overhead compared to existing brute-force approaches, advancing the practical deployment of a diffusion-based image editing framework in the real world. Codes can be found at https://github.com/chaupham1709/AutoEdit.git.

Nikhil Manali, David Doermann, Mahesh Desai

Organizational charts, also known as org charts, are critical representations of an organization's structure and the hierarchical relationships between its components and positions. However, manually extracting information from org charts can be error-prone and time-consuming. To solve this, we present an automated and end-to-end approach that uses computer vision, deep learning, and natural language processing techniques. Additionally, we propose a metric to evaluate the completeness and hierarchical accuracy of the extracted information. This approach has the potential to improve organizational restructuring and resource utilization by providing a clear and concise representation of the organizational structure. Our study lays a foundation for further research on the topic of hierarchical chart analysis.

Jihao Qiu, Yuan Zhang, Xi Tang et al.

Videos carry rich visual information including object description, action, interaction, etc., but the existing multimodal large language models (MLLMs) fell short in referential understanding scenarios such as video-based referring. In this paper, we present Artemis, an MLLM that pushes video-based referential understanding to a finer level. Given a video, Artemis receives a natural-language question with a bounding box in any video frame and describes the referred target in the entire video. The key to achieving this goal lies in extracting compact, target-specific video features, where we set a solid baseline by tracking and selecting spatiotemporal features from the video. We train Artemis on the newly established VideoRef45K dataset with 45K video-QA pairs and design a computationally efficient, three-stage training procedure. Results are promising both quantitatively and qualitatively. Additionally, we show that \model can be integrated with video grounding and text summarization tools to understand more complex scenarios. Code and data are available at https://github.com/qiujihao19/Artemis.

Jay Lal, Aditya Mitkari, Mahesh Bhosale et al.

Data extraction from line-chart images is an essential component of the automated document understanding process, as line charts are a ubiquitous data visualization format. However, the amount of visual and structural variations in multi-line graphs makes them particularly challenging for automated parsing. Existing works, however, are not robust to all these variations, either taking an all-chart unified approach or relying on auxiliary information such as legends for line data extraction. In this work, we propose LineFormer, a robust approach to line data extraction using instance segmentation. We achieve state-of-the-art performance on several benchmark synthetic and real chart datasets. Our implementation is available at https://github.com/TheJaeLal/LineFormer .

Runqi Wang, Baochang Zhang, Li'an Zhuo et al.

Conventional gradient descent methods compute the gradients for multiple variables through the partial derivative. Treating the coupled variables independently while ignoring the interaction, however, leads to an insufficient optimization for bilinear models. In this paper, we propose a dependable learning based on Cogradient Descent (CoGD) algorithm to address the bilinear optimization problem, providing a systematic way to coordinate the gradients of coupling variables based on a kernelized projection function. CoGD is introduced to solve bilinear problems when one variable is with sparsity constraint, as often occurs in modern learning paradigms. CoGD can also be used to decompose the association of features and weights, which further generalizes our method to better train convolutional neural networks (CNNs) and improve the model capacity. CoGD is applied in representative bilinear problems, including image reconstruction, image inpainting, network pruning and CNN training. Extensive experiments show that CoGD improves the state-of-the-arts by significant margins. Code is available at {https://github.com/bczhangbczhang/CoGD}.

Mingyuan Mao, Yuxin Tian, Baochang Zhang et al.

Modern CNN-based object detectors focus on feature configuration during training but often ignore feature optimization during inference. In this paper, we propose a new feature optimization approach to enhance features and suppress background noise in both the training and inference stages. We introduce a generic Inference-aware Feature Filtering (IFF) module that can easily be combined with modern detectors, resulting in our iffDetector. Unlike conventional open-loop feature calculation approaches without feedback, the IFF module performs closed-loop optimization by leveraging high-level semantics to enhance the convolutional features. By applying Fourier transform analysis, we demonstrate that the IFF module acts as a negative feedback that theoretically guarantees the stability of feature learning. IFF can be fused with CNN-based object detectors in a plug-and-play manner with negligible computational cost overhead. Experiments on the PASCAL VOC and MS COCO datasets demonstrate that our iffDetector consistently outperforms state-of-the-art methods by significant margins\footnote{The test code and model are anonymously available in https://github.com/anonymous2020new/iffDetector }.

Pengyu Yan, Mahesh Bhosale, Jay Lal et al.

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.

Xuan Gong, Shanglin Li, Yuxiang Bao et al.

Federated learning (FL) is a machine learning paradigm in which distributed local nodes collaboratively train a central model without sharing individually held private data. Existing FL methods either iteratively share local model parameters or deploy co-distillation. However, the former is highly susceptible to private data leakage, and the latter design relies on the prerequisites of task-relevant real data. Instead, we propose a data-free FL framework based on local-to-central collaborative distillation with direct input and output space exploitation. Our design eliminates any requirement of recursive local parameter exchange or auxiliary task-relevant data to transfer knowledge, thereby giving direct privacy control to local users. In particular, to cope with the inherent data heterogeneity across locals, our technique learns to distill input on which each local model produces consensual yet unique results to represent each expertise. Our proposed FL framework achieves notable privacy-utility trade-offs with extensive experiments on image classification and segmentation tasks under various real-world heterogeneous federated learning settings on both natural and medical images.

Chau Pham, Hoang Phan, David Doermann et al.

The personalization model has gained significant attention in image generation yet remains underexplored for large vision-language models (LVLMs). Beyond generic ones, with personalization, LVLMs handle interactive dialogues using referential concepts (e.g., ``Mike and Susan are talking.'') instead of the generic form (e.g., ``a boy and a girl are talking.''), making the conversation more customizable and referentially friendly. In addition, PLVM is equipped to continuously add new concepts during a dialogue without incurring additional costs, which significantly enhances the practicality. PLVM proposes Aligner, a pre-trained visual encoder to align referential concepts with the queried images. During the dialogues, it extracts features of reference images with these corresponding concepts and recognizes them in the queried image, enabling personalization. We note that the computational cost and parameter count of the Aligner are negligible within the entire framework. With comprehensive qualitative and quantitative analyses, we reveal the effectiveness and superiority of PLVM.

Mahesh Bhosale, Abdul Wasi, Yuanhao Zhai et al.

Diffusion-based generative models have shown promise in synthesizing histopathology images to address data scarcity caused by privacy constraints. Diagnostic text reports provide high-level semantic descriptions, and masks offer fine-grained spatial structures essential for representing distinct morphological regions. However, public datasets lack paired text and mask data for the same histopathological images, limiting their joint use in image generation. This constraint restricts the ability to fully exploit the benefits of combining both modalities for enhanced control over semantics and spatial details. To overcome this, we propose PathDiff, a diffusion framework that effectively learns from unpaired mask-text data by integrating both modalities into a unified conditioning space. PathDiff allows precise control over structural and contextual features, generating high-quality, semantically accurate images. PathDiff also improves image fidelity, text-image alignment, and faithfulness, enhancing data augmentation for downstream tasks like nuclei segmentation and classification. Extensive experiments demonstrate its superiority over existing methods.

Saptarshi Saha, Dhruv Vansraj Rathore, Soumadeep Saha et al.

Quantifying the causal influence of input features within neural networks has become a topic of increasing interest. Existing approaches typically assess direct, indirect, and total causal effects. This work treats NNs as structural causal models (SCMs) and extends our focus to include intrinsic causal contributions (ICC). We propose an identifiable generative post-hoc framework for quantifying ICC. We also draw a relationship between ICC and Sobol' indices. Our experiments on synthetic and real-world datasets demonstrate that ICC generates more intuitive and reliable explanations compared to existing global explanation techniques.

Mohd Faiz Ansari, Rakshit Sandilya, Mohammed Javed et al.

Road anomalies can be defined as irregularities on the road surface or in the surface itself. Some may be intentional (such as speedbumps), accidental (such as materials falling off a truck), or the result of roads' excessive use or low or no maintenance, such as potholes. Despite their varying origins, these irregularities often harm vehicles substantially. Speed bumps are intentionally placed for safety but are dangerous due to their non-standard shape, size, and lack of proper markings. Potholes are unintentional and can also cause severe damage. To address the detection of these anomalies, we need an automated road monitoring system. Today, various systems exist that use visual information to track these anomalies. Still, due to poor lighting conditions and improper or missing markings, they may go undetected and have severe consequences for public transport, automated vehicles, etc. In this paper, the Enhanced Temporal-BiLSTM Network (ETLNet) is introduced as a novel approach that integrates two Temporal Convolutional Network (TCN) layers with a Bidirectional Long Short-Term Memory (BiLSTM) layer. This combination is tailored to detect anomalies effectively irrespective of lighting conditions, as it depends not on visuals but smartphone inertial sensor data. Our methodology employs accelerometer and gyroscope sensors, typically in smartphones, to gather data on road conditions. Empirical evaluations demonstrate that the ETLNet model maintains an F1-score for detecting speed bumps of 99.3%. The ETLNet model's robustness and efficiency significantly advance automated road surface monitoring technologies.

Yuanhao Zhai, Kevin Lin, Zhengyuan Yang et al.

Image diffusion distillation achieves high-fidelity generation with very few sampling steps. However, applying these techniques directly to video diffusion often results in unsatisfactory frame quality due to the limited visual quality in public video datasets. This affects the performance of both teacher and student video diffusion models. Our study aims to improve video diffusion distillation while improving frame appearance using abundant high-quality image data. We propose motion consistency model (MCM), a single-stage video diffusion distillation method that disentangles motion and appearance learning. Specifically, MCM includes a video consistency model that distills motion from the video teacher model, and an image discriminator that enhances frame appearance to match high-quality image data. This combination presents two challenges: (1) conflicting frame learning objectives, as video distillation learns from low-quality video frames while the image discriminator targets high-quality images; and (2) training-inference discrepancies due to the differing quality of video samples used during training and inference. To address these challenges, we introduce disentangled motion distillation and mixed trajectory distillation. The former applies the distillation objective solely to the motion representation, while the latter mitigates training-inference discrepancies by mixing distillation trajectories from both the low- and high-quality video domains. Extensive experiments show that our MCM achieves the state-of-the-art video diffusion distillation performance. Additionally, our method can enhance frame quality in video diffusion models, producing frames with high aesthetic scores or specific styles without corresponding video data.

Yuanhao Zhai, Ziyi Liu, Zhenyu Wu et al.

Deep learning models have a risk of utilizing spurious clues to make predictions, such as recognizing actions based on the background scene. This issue can severely degrade the open-set action recognition performance when the testing samples have different scene distributions from the training samples. To mitigate this problem, we propose a novel method, called Scene-debiasing Open-set Action Recognition (SOAR), which features an adversarial scene reconstruction module and an adaptive adversarial scene classification module. The former prevents the decoder from reconstructing the video background given video features, and thus helps reduce the background information in feature learning. The latter aims to confuse scene type classification given video features, with a specific emphasis on the action foreground, and helps to learn scene-invariant information. In addition, we design an experiment to quantify the scene bias. The results indicate that the current open-set action recognizers are biased toward the scene, and our proposed SOAR method better mitigates such bias. Furthermore, our extensive experiments demonstrate that our method outperforms state-of-the-art methods, and the ablation studies confirm the effectiveness of our proposed modules.

Yuanhao Zhai, Tianyu Luan, David Doermann et al.

As advanced image manipulation techniques emerge, detecting the manipulation becomes increasingly important. Despite the success of recent learning-based approaches for image manipulation detection, they typically require expensive pixel-level annotations to train, while exhibiting degraded performance when testing on images that are differently manipulated compared with training images. To address these limitations, we propose weakly-supervised image manipulation detection, such that only binary image-level labels (authentic or tampered with) are required for training purpose. Such a weakly-supervised setting can leverage more training images and has the potential to adapt quickly to new manipulation techniques. To improve the generalization ability, we propose weakly-supervised self-consistency learning (WSCL) to leverage the weakly annotated images. Specifically, two consistency properties are learned: multi-source consistency (MSC) and inter-patch consistency (IPC). MSC exploits different content-agnostic information and enables cross-source learning via an online pseudo label generation and refinement process. IPC performs global pair-wise patch-patch relationship reasoning to discover a complete region of manipulation. Extensive experiments validate that our WSCL, even though is weakly supervised, exhibits competitive performance compared with fully-supervised counterpart under both in-distribution and out-of-distribution evaluations, as well as reasonable manipulation localization ability.

Pengyu Yan, Saleem Ahmed, David Doermann

As a prerequisite of chart data extraction, the accurate detection of chart basic elements is essential and mandatory. In contrast to object detection in the general image domain, chart element detection relies heavily on context information as charts are highly structured data visualization formats. To address this, we propose a novel method CACHED, which stands for Context-Aware Chart Element Detection, by integrating a local-global context fusion module consisting of visual context enhancement and positional context encoding with the Cascade R-CNN framework. To improve the generalization of our method for broader applicability, we refine the existing chart element categorization and standardized 18 classes for chart basic elements, excluding plot elements. Our CACHED method, with the updated category of chart elements, achieves state-of-the-art performance in our experiments, underscoring the importance of context in chart element detection. Extending our method to the bar plot detection task, we obtain the best result on the PMC test dataset.

Runqi Wang, Xiaoyue Duan, Baochang Zhang et al.

A human's attention can intuitively adapt to corrupted areas of an image by recalling a similar uncorrupted image they have previously seen. This observation motivates us to improve the attention of adversarial images by considering their clean counterparts. To accomplish this, we introduce Associative Adversarial Learning (AAL) into adversarial learning to guide a selective attack. We formulate the intrinsic relationship between attention and attack (perturbation) as a coupling optimization problem to improve their interaction. This leads to an attention backtracking algorithm that can effectively enhance the attention's adversarial robustness. Our method is generic and can be used to address a variety of tasks by simply choosing different kernels for the associative attention that select other regions for a specific attack. Experimental results show that the selective attack improves the model's performance. We show that our method improves the recognition accuracy of adversarial training on ImageNet by 8.32% compared with the baseline. It also increases object detection mAP on PascalVOC by 2.02% and recognition accuracy of few-shot learning on miniImageNet by 1.63%.

Joseph P. Distefano, Hemanth Manjunatha, Souma Chowdhury et al.

Human-swarm interaction has recently gained attention due to its plethora of new applications in disaster relief, surveillance, rescue, and exploration. However, if the task difficulty increases, the performance of the human operator decreases, thereby decreasing the overall efficacy of the human-swarm team. Thus, it is critical to identify the task difficulty and adaptively allocate the task to the human operator to maintain optimal performance. In this direction, we study the classification of task difficulty in a human-swarm interaction experiment performing a target search mission. The human may control platoons of unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs) to search a partially observable environment during the target search mission. The mission complexity is increased by introducing adversarial teams that humans may only see when the environment is explored. While the human is completing the mission, their brain activity is recorded using an electroencephalogram (EEG), which is used to classify the task difficulty. We have used two different approaches for classification: A feature-based approach using coherence values as input and a deep learning-based approach using raw EEG as input. Both approaches can classify the task difficulty well above the chance. The results showed the importance of the occipital lobe (O1 and O2) coherence feature with the other brain regions. Moreover, we also study individual differences (expert vs. novice) in the classification results. The analysis revealed that the temporal lobe in experts (T4 and T3) is predominant for task difficulty classification compared with novices.

Amir Behjat, Hemanth Manjunatha, Prajit KrisshnaKumar et al.

To accomplish complex swarm robotic missions in the real world, one needs to plan and execute a combination of single robot behaviors, group primitives such as task allocation, path planning, and formation control, and mission-specific objectives such as target search and group coverage. Most such missions are designed manually by teams of robotics experts. Recent work in automated approaches to learning swarm behavior has been limited to individual primitives with sparse work on learning complete missions. This paper presents a systematic approach to learn tactical mission-specific policies that compose primitives in a swarm to accomplish the mission efficiently using neural networks with special input and output encoding. To learn swarm tactics in an adversarial environment, we employ a combination of 1) map-to-graph abstraction, 2) input/output encoding via Pareto filtering of points of interest and clustering of robots, and 3) learning via neuroevolution and policy gradient approaches. We illustrate this combination as critical to providing tractable learning, especially given the computational cost of simulating swarm missions of this scale and complexity. Successful mission completion outcomes are demonstrated with up to 60 robots. In addition, a close match in the performance statistics in training and testing scenarios shows the potential generalizability of the proposed framework.

Manan Oza, Sukalpa Chanda, David Doermann

Faces generated using generative adversarial networks (GANs) have reached unprecedented realism. These faces, also known as "Deep Fakes", appear as realistic photographs with very little pixel-level distortions. While some work has enabled the training of models that lead to the generation of specific properties of the subject, generating a facial image based on a natural language description has not been fully explored. For security and criminal identification, the ability to provide a GAN-based system that works like a sketch artist would be incredibly useful. In this paper, we present a novel approach to generate facial images from semantic text descriptions. The learned model is provided with a text description and an outline of the type of face, which the model uses to sketch the features. Our models are trained using an Affine Combination Module (ACM) mechanism to combine the text embedding from BERT and the GAN latent space using a self-attention matrix. This avoids the loss of features due to inadequate "attention", which may happen if text embedding and latent vector are simply concatenated. Our approach is capable of generating images that are very accurately aligned to the exhaustive textual descriptions of faces with many fine detail features of the face and helps in generating better images. The proposed method is also capable of making incremental changes to a previously generated image if it is provided with additional textual descriptions or sentences.

Yuanhao Zhai, Le Wang, David Doermann et al.

This technical report presents our solution to the HACS Temporal Action Localization Challenge 2021, Weakly-Supervised Learning Track. The goal of weakly-supervised temporal action localization is to temporally locate and classify action of interest in untrimmed videos given only video-level labels. We adopt the two-stream consensus network (TSCN) as the main framework in this challenge. The TSCN consists of a two-stream base model training procedure and a pseudo ground truth learning procedure. The base model training encourages the model to predict reliable predictions based on single modality (i.e., RGB or optical flow), based on the fusion of which a pseudo ground truth is generated and in turn used as supervision to train the base models. On the HACS v1.1.1 dataset, without fine-tuning the feature-extraction I3D models, our method achieves 22.20% on the validation set and 21.68% on the testing set in terms of average mAP. Our solution ranked the 2nd in this challenge, and we hope our method can serve as a baseline for future academic research.

Teli Ma, Mingyuan Mao, Honghui Zheng et al.

Object detection with Transformers (DETR) has achieved a competitive performance over traditional detectors, such as Faster R-CNN. However, the potential of DETR remains largely unexplored for the more challenging task of arbitrary-oriented object detection problem. We provide the first attempt and implement Oriented Object DEtection with TRansformer ($\bf O^2DETR$) based on an end-to-end network. The contributions of $\rm O^2DETR$ include: 1) we provide a new insight into oriented object detection, by applying Transformer to directly and efficiently localize objects without a tedious process of rotated anchors as in conventional detectors; 2) we design a simple but highly efficient encoder for Transformer by replacing the attention mechanism with depthwise separable convolution, which can significantly reduce the memory and computational cost of using multi-scale features in the original Transformer; 3) our $\rm O^2DETR$ can be another new benchmark in the field of oriented object detection, which achieves up to 3.85 mAP improvement over Faster R-CNN and RetinaNet. We simply fine-tune the head mounted on $\rm O^2DETR$ in a cascaded architecture and achieve a competitive performance over SOTA in the DOTA dataset.

Mingyuan Mao, Baochang Zhang, David Doermann et al.

Model ensembles are becoming one of the most effective approaches for improving object detection performance already optimized for a single detector. Conventional methods directly fuse bounding boxes but typically fail to consider proposal qualities when combining detectors. This leads to a new problem of confidence discrepancy for the detector ensembles. The confidence has little effect on single detectors but significantly affects detector ensembles. To address this issue, we propose a novel ensemble called the Probabilistic Ranking Aware Ensemble (PRAE) that refines the confidence of bounding boxes from detectors. By simultaneously considering the category and the location on the same validation set, we obtain a more reliable confidence based on statistical probability. We can then rank the detected bounding boxes for assembly. We also introduce a bandit approach to address the confidence imbalance problem caused by the need to deal with different numbers of boxes at different confidence levels. We use our PRAE-based non-maximum suppression (P-NMS) to replace the conventional NMS method in ensemble learning. Experiments on the PASCAL VOC and COCO2017 datasets demonstrate that our PRAE method consistently outperforms state-of-the-art methods by significant margins.

Leighton Collins, Payam Ghassemi, Ehsan T. Esfahani et al.

This paper presents a novel multi-robot coverage path planning (CPP) algorithm - aka SCoPP - that provides a time-efficient solution, with workload balanced plans for each robot in a multi-robot system, based on their initial states. This algorithm accounts for discontinuities (e.g., no-fly zones) in a specified area of interest, and provides an optimized ordered list of way-points per robot using a discrete, computationally efficient, nearest neighbor path planning algorithm. This algorithm involves five main stages, which include the transformation of the user's input as a set of vertices in geographical coordinates, discretization, load-balanced partitioning, auctioning of conflict cells in a discretized space, and a path planning procedure. To evaluate the effectiveness of the primary algorithm, a multi-unmanned aerial vehicle (UAV) post-flood assessment application is considered, and the performance of the algorithm is tested on three test maps of varying sizes. Additionally, our method is compared with a state-of-the-art method created by Guasella et al. Further analyses on scalability and computational time of SCoPP are conducted. The results show that SCoPP is superior in terms of mission completion time; its computing time is found to be under 2 mins for a large map covered by a 150-robot team, thereby demonstrating its computationally scalability.

Huan Chang, Yicheng Chen, Baochang Zhang et al.

Unmanned Aerial vehicles (UAVs) are widely used as network processors in mobile networks, but more recently, UAVs have been used in Mobile Edge Computing as mobile servers. However, there are significant challenges to use UAVs in complex environments with obstacles and cooperation between UAVs. We introduce a new multi-UAV Mobile Edge Computing platform, which aims to provide better Quality-of-Service and path planning based on reinforcement learning to address these issues. The contributions of our work include: 1) optimizing the quality of service for mobile edge computing and path planning in the same reinforcement learning framework; 2) using a sigmoid-like function to depict the terminal users' demand to ensure a higher quality of service; 3) applying synthetic considerations of the terminal users' demand, risk and geometric distance in reinforcement learning reward matrix to ensure the quality of service, risk avoidance, and the cost-savings. Simulations have shown the effectiveness and feasibility of our platform, which can help advance related researches.

Xuan Gong, Xin Xia, Wentao Zhu et al.

In recent years, deep learning has dominated progress in the field of medical image analysis. We find however, that the ability of current deep learning approaches to represent the complex geometric structures of many medical images is insufficient. One limitation is that deep learning models require a tremendous amount of data, and it is very difficult to obtain a sufficient amount with the necessary detail. A second limitation is that there are underlying features of these medical images that are well established, but the black-box nature of existing convolutional neural networks (CNNs) do not allow us to exploit them. In this paper, we revisit Gabor filters and introduce a deformable Gabor convolution (DGConv) to expand deep networks interpretability and enable complex spatial variations. The features are learned at deformable sampling locations with adaptive Gabor convolutions to improve representativeness and robustness to complex objects. The DGConv replaces standard convolutional layers and is easily trained end-to-end, resulting in deformable Gabor feature network (DGFN) with few additional parameters and minimal additional training cost. We introduce DGFN for addressing deep multi-instance multi-label classification on the INbreast dataset for mammograms and on the ChestX-ray14 dataset for pulmonary x-ray images.

Wenyu Zhao, Teli Ma, Xuan Gong et al.

Edge computing is promising to become one of the next hottest topics in artificial intelligence because it benefits various evolving domains such as real-time unmanned aerial systems, industrial applications, and the demand for privacy protection. This paper reviews recent advances on binary neural network (BNN) and 1-bit CNN technologies that are well suitable for front-end, edge-based computing. We introduce and summarize existing work and classify them based on gradient approximation, quantization, architecture, loss functions, optimization method, and binary neural architecture search. We also introduce applications in the areas of computer vision and speech recognition and discuss future applications for edge computing.

Hanlin Chen, Li'an Zhuo, Baochang Zhang et al.

Traditional neural architecture search (NAS) has a significant impact in computer vision by automatically designing network architectures for various tasks. In this paper, binarized neural architecture search (BNAS), with a search space of binarized convolutions, is introduced to produce extremely compressed models to reduce huge computational cost on embedded devices for edge computing. The BNAS calculation is more challenging than NAS due to the learning inefficiency caused by optimization requirements and the huge architecture space, and the performance loss when handling the wild data in various computing applications. To address these issues, we introduce operation space reduction and channel sampling into BNAS to significantly reduce the cost of searching. This is accomplished through a performance-based strategy that is robust to wild data, which is further used to abandon less potential operations. Furthermore, we introduce the Upper Confidence Bound (UCB) to solve 1-bit BNAS. Two optimization methods for binarized neural networks are used to validate the effectiveness of our BNAS. Extensive experiments demonstrate that the proposed BNAS achieves a comparable performance to NAS on both CIFAR and ImageNet databases. An accuracy of $96.53\%$ vs. $97.22\%$ is achieved on the CIFAR-10 dataset, but with a significantly compressed model, and a $40\%$ faster search than the state-of-the-art PC-DARTS. On the wild face recognition task, our binarized models achieve a performance similar to their corresponding full-precision models.

Hanlin Chen, Baochang Zhang, Song Xue et al.

Deep convolutional neural networks (DCNNs) have dominated as the best performers in machine learning, but can be challenged by adversarial attacks. In this paper, we defend against adversarial attacks using neural architecture search (NAS) which is based on a comprehensive search of denoising blocks, weight-free operations, Gabor filters and convolutions. The resulting anti-bandit NAS (ABanditNAS) incorporates a new operation evaluation measure and search process based on the lower and upper confidence bounds (LCB and UCB). Unlike the conventional bandit algorithm using UCB for evaluation only, we use UCB to abandon arms for search efficiency and LCB for a fair competition between arms. Extensive experiments demonstrate that ABanditNAS is faster than other NAS methods, while achieving an $8.73\%$ improvement over prior arts on CIFAR-10 under PGD-$7$.

Li'an Zhuo, Baochang Zhang, Linlin Yang et al.

Conventional learning methods simplify the bilinear model by regarding two intrinsically coupled factors independently, which degrades the optimization procedure. One reason lies in the insufficient training due to the asynchronous gradient descent, which results in vanishing gradients for the coupled variables. In this paper, we introduce a Cogradient Descent algorithm (CoGD) to address the bilinear problem, based on a theoretical framework to coordinate the gradient of hidden variables via a projection function. We solve one variable by considering its coupling relationship with the other, leading to a synchronous gradient descent to facilitate the optimization procedure. Our algorithm is applied to solve problems with one variable under the sparsity constraint, which is widely used in the learning paradigm. We validate our CoGD considering an extensive set of applications including image reconstruction, inpainting, and network pruning. Experiments show that it improves the state-of-the-art by a significant margin.

Li'an Zhuo, Baochang Zhang, Hanlin Chen et al.

Neural architecture search (NAS) proves to be among the best approaches for many tasks by generating an application-adaptive neural architecture, which is still challenged by high computational cost and memory consumption. At the same time, 1-bit convolutional neural networks (CNNs) with binarized weights and activations show their potential for resource-limited embedded devices. One natural approach is to use 1-bit CNNs to reduce the computation and memory cost of NAS by taking advantage of the strengths of each in a unified framework. To this end, a Child-Parent (CP) model is introduced to a differentiable NAS to search the binarized architecture (Child) under the supervision of a full-precision model (Parent). In the search stage, the Child-Parent model uses an indicator generated by the child and parent model accuracy to evaluate the performance and abandon operations with less potential. In the training stage, a kernel-level CP loss is introduced to optimize the binarized network. Extensive experiments demonstrate that the proposed CP-NAS achieves a comparable accuracy with traditional NAS on both the CIFAR and ImageNet databases. It achieves the accuracy of $95.27\%$ on CIFAR-10, $64.3\%$ on ImageNet with binarized weights and activations, and a $30\%$ faster search than prior arts.