Xu An Wang

Beilei Cui, Minqing Zhang, Mengya Xu et al.

Noisy label problems are inevitably in existence within medical image segmentation causing severe performance degradation. Previous segmentation methods for noisy label problems only utilize a single image while the potential of leveraging the correlation between images has been overlooked. Especially for video segmentation, adjacent frames contain rich contextual information beneficial in cognizing noisy labels. Based on two insights, we propose a Multi-Scale Temporal Feature Affinity Learning (MS-TFAL) framework to resolve noisy-labeled medical video segmentation issues. First, we argue the sequential prior of videos is an effective reference, i.e., pixel-level features from adjacent frames are close in distance for the same class and far in distance otherwise. Therefore, Temporal Feature Affinity Learning (TFAL) is devised to indicate possible noisy labels by evaluating the affinity between pixels in two adjacent frames. We also notice that the noise distribution exhibits considerable variations across video, image, and pixel levels. In this way, we introduce Multi-Scale Supervision (MSS) to supervise the network from three different perspectives by re-weighting and refining the samples. This design enables the network to concentrate on clean samples in a coarse-to-fine manner. Experiments with both synthetic and real-world label noise demonstrate that our method outperforms recent state-of-the-art robust segmentation approaches. Code is available at https://github.com/BeileiCui/MS-TFAL.

Kaixiang Yang, Wenqi Shan, Xudong Li et al.

Multi-modal brain tumor segmentation typically involves four magnetic resonance imaging (MRI) modalities, while incomplete modalities significantly degrade performance. Existing solutions employ explicit or implicit modality adaptation, aligning features across modalities or learning a fused feature robust to modality incompleteness. They share a common goal of encouraging each modality to express both itself and the others. However, the two expression abilities are entangled as a whole in a seamless feature space, resulting in prohibitive learning burdens. In this paper, we propose DeMoSeg to enhance the modality adaptation by Decoupling the task of representing the ego and other Modalities for robust incomplete multi-modal Segmentation. The decoupling is super lightweight by simply using two convolutions to map each modality onto four feature sub-spaces. The first sub-space expresses itself (Self-feature), while the remaining sub-spaces substitute for other modalities (Mutual-features). The Self- and Mutual-features interactively guide each other through a carefully-designed Channel-wised Sparse Self-Attention (CSSA). After that, a Radiologist-mimic Cross-modality expression Relationships (RCR) is introduced to have available modalities provide Self-feature and also `lend' their Mutual-features to compensate for the absent ones by exploiting the clinical prior knowledge. The benchmark results on BraTS2020, BraTS2018 and BraTS2015 verify the DeMoSeg's superiority thanks to the alleviated modality adaptation difficulty. Concretely, for BraTS2020, DeMoSeg increases Dice by at least 0.92%, 2.95% and 4.95% on whole tumor, tumor core and enhanced tumor regions, respectively, compared to other state-of-the-arts. Codes are at https://github.com/kk42yy/DeMoSeg

Mengya Xu, Ziqi Guo, An Wang et al.

As a crucial and intricate task in robotic minimally invasive surgery, reconstructing surgical scenes using stereo or monocular endoscopic video holds immense potential for clinical applications. NeRF-based techniques have recently garnered attention for the ability to reconstruct scenes implicitly. On the other hand, Gaussian splatting-based 3D-GS represents scenes explicitly using 3D Gaussians and projects them onto a 2D plane as a replacement for the complex volume rendering in NeRF. However, these methods face challenges regarding surgical scene reconstruction, such as slow inference, dynamic scenes, and surgical tool occlusion. This work explores and reviews state-of-the-art (SOTA) approaches, discussing their innovations and implementation principles. Furthermore, we replicate the models and conduct testing and evaluation on two datasets. The test results demonstrate that with advancements in these techniques, achieving real-time, high-quality reconstructions becomes feasible.

Kangneng Zhou, Daiheng Gao, Xuan Wang et al.

3D-aware portrait editing has a wide range of applications in multiple fields. However, current approaches are limited due that they can only perform mask-guided or text-based editing. Even by fusing the two procedures into a model, the editing quality and stability cannot be ensured. To address this limitation, we propose \textbf{MaTe3D}: mask-guided text-based 3D-aware portrait editing. In this framework, first, we introduce a new SDF-based 3D generator which learns local and global representations with proposed SDF and density consistency losses. This enhances masked-based editing in local areas; second, we present a novel distillation strategy: Conditional Distillation on Geometry and Texture (CDGT). Compared to exiting distillation strategies, it mitigates visual ambiguity and avoids mismatch between texture and geometry, thereby producing stable texture and convincing geometry while editing. Additionally, we create the CatMask-HQ dataset, a large-scale high-resolution cat face annotation for exploration of model generalization and expansion. We perform expensive experiments on both the FFHQ and CatMask-HQ datasets to demonstrate the editing quality and stability of the proposed method. Our method faithfully generates a 3D-aware edited face image based on a modified mask and a text prompt. Our code and models will be publicly released.

Xiaowen Zhang, Zhenyu Bi, Patrick Lachance et al.

As cosmological simulations and their associated software become increasingly complex, physicists face the challenge of searching through vast amounts of literature and user manuals to extract simulation parameters from dense academic papers, each using different models and formats. Translating these parameters into executable scripts remains a time-consuming and error-prone process. To improve efficiency in physics research and accelerate the cosmological simulation process, we introduce SimAgents, a multi-agent system designed to automate both parameter configuration from the literature and preliminary analysis for cosmology research. SimAgents is powered by specialized LLM agents capable of physics reasoning, simulation software validation, and tool execution. These agents collaborate through structured communication, ensuring that extracted parameters are physically meaningful, internally consistent, and software-compliant. We also construct a cosmological parameter extraction evaluation dataset by collecting over 40 simulations in published papers from Arxiv and leading journals that cover diverse simulation types. Experiments on the dataset demonstrate a strong performance of SimAgents, highlighting its effectiveness and potential to accelerate scientific research for physicists. Our demonstration video is available at: https://youtu.be/w1zLpm_CaWA. The complete system and dataset are publicly available at https://github.com/xwzhang98/SimAgents.

Weichao Pan, Bohan Xu, Xu Wang et al.

Fire detection in dynamic environments faces continuous challenges, including the interference of illumination changes, many false detections or missed detections, and it is difficult to achieve both efficiency and accuracy. To address the problem of feature extraction limitation and information loss in the existing YOLO-based models, this study propose You Only Look Once for Fire Detection with Attention-guided Inverted Residual and Dual-pooling Downscale Fusion (YOLO-FireAD) with two core innovations: (1) Attention-guided Inverted Residual Block (AIR) integrates hybrid channel-spatial attention with inverted residuals to adaptively enhance fire features and suppress environmental noise; (2) Dual Pool Downscale Fusion Block (DPDF) preserves multi-scale fire patterns through learnable fusion of max-average pooling outputs, mitigating small-fire detection failures. Extensive evaluation on two public datasets shows the efficient performance of our model. Our proposed model keeps the sum amount of parameters (1.45M, 51.8% lower than YOLOv8n) (4.6G, 43.2% lower than YOLOv8n), and mAP75 is higher than the mainstream real-time object detection models YOLOv8n, YOL-Ov9t, YOLOv10n, YOLO11n, YOLOv12n and other YOLOv8 variants 1.3-5.5%. For more details, please visit our repository: https://github.com/JEFfersusu/YOLO-FireAD

Rulin Zhou, Wenlong He, An Wang et al.

Accurate point tracking in surgical environments remains challenging due to complex visual conditions, including smoke occlusion, specular reflections, and tissue deformation. While existing surgical tracking datasets provide coordinate information, they lack the semantic context necessary to understand tracking failure mechanisms. We introduce VL-SurgPT, the first large-scale multimodal dataset that bridges visual tracking with textual descriptions of point status in surgical scenes. The dataset comprises 908 in vivo video clips, including 754 for tissue tracking (17,171 annotated points across five challenging scenarios) and 154 for instrument tracking (covering seven instrument types with detailed keypoint annotations). We establish comprehensive benchmarks using eight state-of-the-art tracking methods and propose TG-SurgPT, a text-guided tracking approach that leverages semantic descriptions to improve robustness in visually challenging conditions. Experimental results demonstrate that incorporating point status information significantly improves tracking accuracy and reliability, particularly in adverse visual scenarios where conventional vision-only methods struggle. By bridging visual and linguistic modalities, VL-SurgPT enables the development of context-aware tracking systems crucial for advancing computer-assisted surgery applications that can maintain performance even under challenging intraoperative conditions.

Xingwu Sun, Shuaipeng Li, Ruobing Xie et al.

Low-precision training is considered an effective strategy for reducing both training and downstream inference costs. Previous scaling laws for precision mainly focus on integer quantization, which pay less attention to the constituents in floating-point (FP) quantization, and thus cannot well fit the LLM losses in this scenario. In contrast, while FP quantization training is more commonly implemented in production, it's research has been relatively superficial. In this paper, we thoroughly explore the effects of FP quantization targets, exponent bits, mantissa bits, and the calculation granularity of the scaling factor in FP quantization training performance of LLM models. In addition to an accurate FP quantization unified scaling law, we also provide valuable suggestions for the community: (1) Exponent bits contribute slightly more to the model performance than mantissa bits. We provide the optimal exponent-mantissa bit ratio for different bit numbers, which is available for future reference by hardware manufacturers; (2) We discover the formation of the critical data size in low-precision LLM training. Too much training data exceeding the critical data size will inversely bring in degradation of LLM performance; (3) The optimal FP quantization precision is directly proportional to the computational power, but within a wide computational power range. We estimate that the best cost-performance precision should lie between 4-8 bits.

Xiao Liu, Mingyuan Li, Xu Wang et al.

Unlearning in Federated Learning (FL) presents significant challenges, as models grow and evolve with complex inheritance relationships. This complexity is amplified when blockchain is employed to ensure the integrity and traceability of FL, where the need to edit multiple interlinked blockchain records and update all inherited models complicates the process.In this paper, we introduce Blockchained Federated Unlearning (BlockFUL), a novel framework with a dual-chain structure comprising a live chain and an archive chain for enabling unlearning capabilities within Blockchained FL. BlockFUL introduces two new unlearning paradigms, i.e., parallel and sequential paradigms, which can be effectively implemented through gradient-ascent-based and re-training-based unlearning methods. These methods enhance the unlearning process across multiple inherited models by enabling efficient consensus operations and reducing computational costs. Our extensive experiments validate that these methods effectively reduce data dependency and operational overhead, thereby boosting the overall performance of unlearning inherited models within BlockFUL on CIFAR-10 and Fashion-MNIST datasets using AlexNet, ResNet18, and MobileNetV2 models.

Yi-Fan Zhang, Haihua Yang, Huanyu Zhang et al.

The rapid advancement of Multimodal Large Language Models (MLLMs) has made aligning them with human preferences a critical challenge. Reward Models (RMs) are a core technology for achieving this goal, but a systematic guide for building state-of-the-art Multimodal Reward Models (MRMs) is currently lacking in both academia and industry. Through exhaustive experimental analysis, this paper aims to provide a clear ``recipe'' for constructing high-performance MRMs. We systematically investigate every crucial component in the MRM development pipeline, including \textit{reward modeling paradigms} (e.g., Naive-RM, Critic-based RM, and Generative RM), \textit{reward head architecture}, \textit{training strategies}, \textit{data curation} (covering over ten multimodal and text-only preference datasets), \textit{backbone model} and \textit{model scale}, and \textit{ensemble methods}. Based on these experimental insights, we introduce \textbf{BaseReward}, a powerful and efficient baseline for multimodal reward modeling. BaseReward adopts a simple yet effective architecture, built upon a {Qwen2.5-VL} backbone, featuring an optimized two-layer reward head, and is trained on a carefully curated mixture of high-quality multimodal and text-only preference data. Our results show that BaseReward establishes a new SOTA on major benchmarks such as MM-RLHF-Reward Bench, VL-Reward Bench, and Multimodal Reward Bench, outperforming previous models. Furthermore, to validate its practical utility beyond static benchmarks, we integrate BaseReward into a real-world reinforcement learning pipeline, successfully enhancing an MLLM's performance across various perception, reasoning, and conversational tasks. This work not only delivers a top-tier MRM but, more importantly, provides the community with a clear, empirically-backed guide for developing robust reward models for the next generation of MLLMs.

Rui Tang, Haochen Yin, Guankun Wang et al.

Surgical phase recognition plays a critical role in developing intelligent assistance systems for minimally invasive procedures such as Endoscopic Submucosal Dissection (ESD). However, the high visual similarity across different phases and the lack of structural cues in RGB images pose significant challenges. Depth information offers valuable geometric cues that can complement appearance features by providing insights into spatial relationships and anatomical structures. In this paper, we pioneer the use of depth information for surgical phase recognition and propose Geo-RepNet, a geometry-aware convolutional framework that integrates RGB image and depth information to enhance recognition performance in complex surgical scenes. Built upon a re-parameterizable RepVGG backbone, Geo-RepNet incorporates the Depth-Guided Geometric Prior Generation (DGPG) module that extracts geometry priors from raw depth maps, and the Geometry-Enhanced Multi-scale Attention (GEMA) to inject spatial guidance through geometry-aware cross-attention and efficient multi-scale aggregation. To evaluate the effectiveness of our approach, we construct a nine-phase ESD dataset with dense frame-level annotations from real-world ESD videos. Extensive experiments on the proposed dataset demonstrate that Geo-RepNet achieves state-of-the-art performance while maintaining robustness and high computational efficiency under complex and low-texture surgical environments.

Rulin Zhou, Wenlong He, An Wang et al.

Accurate tissue point tracking in endoscopic videos is critical for robotic-assisted surgical navigation and scene understanding, but remains challenging due to complex deformations, instrument occlusion, and the scarcity of dense trajectory annotations. Existing methods struggle with long-term tracking under these conditions due to limited feature utilization and annotation dependence. We present Endo-TTAP, a novel framework addressing these challenges through: (1) A Multi-Facet Guided Attention (MFGA) module that synergizes multi-scale flow dynamics, DINOv2 semantic embeddings, and explicit motion patterns to jointly predict point positions with uncertainty and occlusion awareness; (2) A two-stage curriculum learning strategy employing an Auxiliary Curriculum Adapter (ACA) for progressive initialization and hybrid supervision. Stage I utilizes synthetic data with optical flow ground truth for uncertainty-occlusion regularization, while Stage II combines unsupervised flow consistency and semi-supervised learning with refined pseudo-labels from off-the-shelf trackers. Extensive validation on two MICCAI Challenge datasets and our collected dataset demonstrates that Endo-TTAP achieves state-of-the-art performance in tissue point tracking, particularly in scenarios characterized by complex endoscopic conditions. The source code and dataset will be available at https://anonymous.4open.science/r/Endo-TTAP-36E5.

Beilei Cui, Long Bai, Mobarakol Islam et al.

Accurate 3D scene reconstruction is essential for numerous medical tasks. Given the challenges in obtaining ground truth data, there has been an increasing focus on self-supervised learning (SSL) for endoscopic depth estimation as a basis for scene reconstruction. While foundation models have shown remarkable progress in visual tasks, their direct application to the medical domain often leads to suboptimal results. However, the visual features from these models can still enhance endoscopic tasks, emphasizing the need for efficient adaptation strategies, which still lack exploration currently. In this paper, we introduce Endo3DAC, a unified framework for endoscopic scene reconstruction that efficiently adapts foundation models. We design an integrated network capable of simultaneously estimating depth maps, relative poses, and camera intrinsic parameters. By freezing the backbone foundation model and training only the specially designed Gated Dynamic Vector-Based Low-Rank Adaptation (GDV-LoRA) with separate decoder heads, Endo3DAC achieves superior depth and pose estimation while maintaining training efficiency. Additionally, we propose a 3D scene reconstruction pipeline that optimizes depth maps' scales, shifts, and a few parameters based on our integrated network. Extensive experiments across four endoscopic datasets demonstrate that Endo3DAC significantly outperforms other state-of-the-art methods while requiring fewer trainable parameters. To our knowledge, we are the first to utilize a single network that only requires surgical videos to perform both SSL depth estimation and scene reconstruction tasks. The code will be released upon acceptance.

Haonan Shi, Tu Ouyang, An Wang

Federated Distillation (FD) has emerged as a popular federated training framework, enabling clients to collaboratively train models without sharing private data. Public Dataset-Assisted Federated Distillation (PDA-FD), which leverages public datasets for knowledge sharing, has become widely adopted. Although PDA-FD enhances privacy compared to traditional Federated Learning, we demonstrate that the use of public datasets still poses significant privacy risks to clients' private training data. This paper presents the first comprehensive privacy analysis of PDA-FD in presence of an honest-but-curious server. We show that the server can exploit clients' inference results on public datasets to extract two critical types of private information: label distributions and membership information of the private training dataset. To quantify these vulnerabilities, we introduce two novel attacks specifically designed for the PDA-FD setting: a label distribution inference attack and innovative membership inference methods based on Likelihood Ratio Attack (LiRA). Through extensive evaluation of three representative PDA-FD frameworks (FedMD, DS-FL, and Cronus), our attacks achieve state-of-the-art performance, with label distribution attacks reaching minimal KL-divergence and membership inference attacks maintaining high True Positive Rates under low False Positive Rate constraints. Our findings reveal significant privacy risks in current PDA-FD frameworks and emphasize the need for more robust privacy protection mechanisms in collaborative learning systems.

Xuan Wang, Zeshan Pang, Yuliang Lu et al.

To provide a foundation for the research of deep learning models, the construction of model pool is an essential step. This paper proposes a Training-Free and Efficient Model Generation and Enhancement Scheme (MGE). This scheme primarily considers two aspects during the model generation process: the distribution of model parameters and model performance. Experiments result shows that generated models are comparable to models obtained through normal training, and even superior in some cases. Moreover, the time consumed in generating models accounts for only 1\% of the time required for normal model training. More importantly, with the enhancement of Evolution-MGE, generated models exhibits competitive generalization ability in few-shot tasks. And the behavioral dissimilarity of generated models has the potential of adversarial defense.

An Wang, Ao Liu, Hieu Hanh Le et al.

Entity-relation extraction aims to jointly solve named entity recognition (NER) and relation extraction (RE). Recent approaches use either one-way sequential information propagation in a pipeline manner or two-way implicit interaction with a shared encoder. However, they still suffer from poor information interaction due to the gap between the different task forms of NER and RE, raising a controversial question whether RE is really beneficial to NER. Motivated by this, we propose a novel and unified cascade framework that combines the advantages of both sequential information propagation and implicit interaction. Meanwhile, it eliminates the gap between the two tasks by reformulating entity-relation extraction as unified span-extraction tasks. Specifically, we propose a selection recurrent network as a shared encoder to encode task-specific independent and shared representations and design two sequential information propagation strategies to realize the sequential information flow between NER and RE. Extensive experiments demonstrate that our approaches can achieve state-of-the-art results on two common benchmarks, ACE05 and SciERC, and effectively model the multi-task interaction, which realizes significant mutual benefits of NER and RE.