paperlist.ai

AI 顶会论文

AI 顶会论文库

检索 30+ 个顶会的论文，包括 NeurIPS、ICLR、ICML、CVPR、ACL、AAAI 等。输入关键词筛选，打开任意一篇查看摘要、作者和链接。

arXiv

Learning Deep Modality-Shared Self-Expressiveness for Image Clustering with Textual Information

Xianghan Meng, Wei He, Zhiyuan Huang, Chun-Guang Li

Leveraging textual information for image clustering has emerged as a promising direction, largely owing to the powerful representations learned by Vision-Language Models (VLMs). Existing approaches typically retrieve a textual counterpart for each image and then refine multimodal representations by directly enforcing cross-modal agreement, e.g., maximizing image-text similarity inherited from pretrained VLMs. However, such a strategy aligns heterogeneous representations across modalities without explicitly modeling the intrinsic structure within each modality and thus might yield unreliable alignment or distort modality-specific structures that are crucial for clustering. In this paper, we propose a simple but principled approach, termed deep modality-shared self-expressive model (DeepMORSE), which discovers cross-modal structures via a modality-shared self-expressive model and simultaneously learns structured representations that conform to a union of modality-specific subspaces. Moreover, we theoretically justify that the modality-shared self-expressive coefficients suppress inter-class noise towards a subspace-preserving solution, and show that mini-batch optimization procedure introduces an implicit regularization onto the self-expressive model. We evaluate our DeepMORSE on six widely used image clustering benchmarks and observe performance improvements exceeding 3% on the UCF-101, DTD-47, and ImageNet-Dogs datasets. In addition, we demonstrate the strong transferability of the learned representations by achieving state-of-the-art performance on downstream tasks such as image retrieval and zero-shot classification---without requiring any task-specific losses or post-processing. The code is available at: https://github.com/mengxianghan123/DeepMORSE.

arXiv

Agentic AI-powered flexible fiber-bundle endoscopy for high-resolution NIR-II fluorescence imaging in vivo

Yanzhao Shi, Yuanhua Liu, Sixin Xu, Wayne Jason Li, Yuyuan Chen, Danyang Xu, Zhisheng Wu, Hanze Yu, Ian Yu-Hong Wong, Simon Ying-Kit Law, Hongjie Dai, Liangqiong Qu, Feifei Wang

Fiber-bundle endoscopy offers a compact and flexible route for clinical fluorescence imaging through natural human orifices, but since its first report in the 1950s, it has remained limited by low spatial resolution, honeycomb artifacts, and inter-core crosstalk. The crosstalk becomes more pronounced at near-infrared-II wavelengths (NIR-II, 1000-3000 nm), a spectral window that offers superior contrast, resolution, and tissue penetration depth for biomedical imaging. Here, we present an AI-powered flexible endoscopy platform that overcomes these constraints through optical-computational co-design: optimizing ultrathin fiber bundles to mitigate crosstalk-induced image blur and enable high-fidelity image transmission across the visible-to-NIR-II spectral range, and developing an Agent-Guided Mixture-of-Experts (GAME) pipeline for honeycomb-artifact removal and image restoration. GAME provides a single restoration entry point for diverse biomedical images acquired with our endoscope, spanning cell, mouse and human samples. It dynamically routes each input to suitable restoration experts via a vision-language model, facilitating image reconstruction with a fourfold resolution improvement beyond the NyquistShannon sampling limit. The utility of our endoscope is demonstrated through in vivo NIR-II imaging of anatomical structures in mice, as well as imaging of the digital micromirror device (DMD)-projected human gastric tube and lymphatic system, paving the way for future clinical translation.

arXiv

Anatomically Consistent Cross-Contrast Super-Resolution of Anisotropic Brain T2w MRI

Mengqi Shen, Haicheng Wang, Meghna Trivedi, Tony J. Wang, Yuanguang Xu, Yingyan Zeng, Yading Yuan

T2-weighted (T2w) brain MRI provides fluid-sensitive soft-tissue contrast that is important for neuro-oncology and radiotherapy planning. However, T2w scans are acquired with anisotropic voxels and appear blurred or stair-stepped on coronal and sagittal views, which obscures small structures and weakens any downstream 3D analysis. We propose VIPP-SR (View-Independent Patched Projection Super-Resolution), a cross-contrast guided super-resolution framework that restores the inter-plane resolution of an existing anisotropic T2w volume without an isotropic ground-truth T2w. VIPP-SR first trains a view-independent patched generator (VIP-GAN) to learn local T1c-to-T2w anatomical correspondence from high-resolution axial slices. The trained generator is then applied to axial, coronal, and sagittal views of the T1c volume to generate three orthogonal T2w estimates. Shape-preserving patching and deepest-skip removal reduce view-specific shortcuts, thereby constraining the generator to learn patch-local representations and enabling the zero-shot inter-plane transfer. Central to VIPP-SR, a projection-based optimization then enforces anatomical consistency across the three view-specific volumes, fusing them by balancing inter-plane self-consistency against per-view data fidelity. The generator is trained on BraTS-MET and evaluated on both the held-out BraTS-MET testing set and the BraTS-GLI cohort without retraining, assessing the cross-cohort generalizability. The results validate that VIPP-SR improves downstream segmentation over the real anisotropic T2w baseline, raising mean-label Dice from 0.330 to 0.465 on BraTS-MET and, zero-shot, from 0.473 to 0.563 on BraTS-GLI and ablation studies identify inter-plane self-consistency as the main source of the gain.

arXiv

DoRF++: Spherical Representation Learning over Doppler Radiance Fields for Robust Wi-Fi Sensing

Navid Hasanzadeh, Shahrokh Valaee

Motivated by the IEEE 802.11bf effort to standardize advanced WLAN sensing, interest in Wi-Fi Channel State Information (CSI) for passive, device-free, and privacy-preserving activity and gesture recognition has grown rapidly. Recent studies have shown that Doppler velocity projections extracted from CSI, which directly reflect human-motion velocity, enable more robust human activity recognition (HAR) and stronger generalization across users and unseen conditions. Nevertheless, reliable generalization under real-world variability remains a major challenge, hindering the adoption of Wi-Fi sensing in real-world applications. To address this challenge, we introduce Doppler Radiance Fields (DoRF), bringing the concept of neural radiance fields (NeRF) from computer vision into Wi-Fi sensing. DoRF models Doppler velocity projections extracted from Wi-Fi CSI as sparse and diverse virtual-camera views of human motion. It then infers a latent 3D motion sequence whose projections along learned effective Doppler directions explain the CSI-derived Doppler observations. The recovered motion is subsequently projected onto an equiangular grid of directions on the unit sphere, producing a spherical representation of the underlying motion. Since DoRF naturally defines the Doppler representation on spheres, we further introduce DoRF++, a spherical-learning design that applies spherical Transformers for activity classification. Experiments on our collected hand-gesture dataset show that DoRF++ significantly outperforms state-of-the-art Wi-Fi-based HAR methods in cross-user generalization accuracy, especially for difficult gestures in settings with a single multi-antenna receiver access point (AP).

arXiv

Teaching Foundation Models to Read mmWave: Pose-Guided Kinematic Representation for Human Behavior Understanding

Duo Zhang, Zhehui Yin, Zhiyun Yao, Haotong Qin, Xusheng Zhang, Hongliu Yang, Jianyu Sun, Junzhe Wang, Zizhou Fan, Michele Magno, Daqing Zhang

Large language model agents need to perceive human behavior in physical environments. Millimeter-wave (mmWave) radar provides a privacy-friendly and contactless sensing modality, but radar observations are difficult to align with language. Existing radar-language methods often rely on synthetic data or lack explicit supervision for human body structure and motion. We present mmMind, a radar-language model that uses synchronized 3D pose as training-only supervision. A spatio-temporal radar encoder is pretrained to capture body configuration and motion dynamics, after which the pose head is removed so that inference requires radar alone. The learned radar representations are then aligned with an LLM for behavior captioning and spatio-temporal question answering. We also introduce mmMind-Bench, a real-world mmWave-language benchmark containing 17.9 hours of recordings from 23 participants across seven indoor environments. Experiments on captioning, question answering, and unseen-action generalization show that mmMind consistently outperforms existing radar-language baselines, while ablations confirm the importance of pose-guided pretraining.

arXiv

Keep the Needle, Prune the Haystack: Defect-Preserving Token Pruning for Efficient Zero-Shot Anomaly Detection

Yanning Hou, Jingyuan Zhang, Xiaoyun Wang, Qixiang Ma, sihang zhou, Ke Xu

Zero-shot visual anomaly detection has achieved remarkable progress, with recent vision-only approaches further improving performance while simplifying the inference pipeline. However, existing methods typically perform dense computation over all images and spatial tokens, despite the fact that normal samples dominate real-world scenarios and anomalies usually occupy only small regions. Token pruning offers a promising solution, but introduces an asymmetric pruning risk in anomaly detection: retaining normal tokens mainly incurs redundant computation, whereas removing anomalous tokens may eliminate the only evidence for detection and localization. This risk is particularly severe in early layers, where pruning provides the greatest computational benefit but anomaly semantics remain unreliable. We propose KeepAD, a defect-preserving token pruning framework that formulates token selection as high-recall, anomaly-aware routing. In shallow layers, KeepAD combines coverage-preserving selection over local $2\times2$ patch neighborhoods with deterministic anomaly rescue to reduce the risk of discarding subtle defects. In deeper layers, frozen normal and abnormal prototypes guide pruning under an image-adaptive token budget, aggressively removing low-risk normal tokens while preserving local anomaly evidence. Dense-to-sparse self-distillation further supervises early token routing without introducing additional inference overhead. Experiments on six industrial and seven medical zero-shot anomaly detection benchmarks show that KeepAD reduces the token retention ratio to below $20\%$, while limiting the average degradation in image-level and pixel-level AUROC to within $2.7$ percentage points. At the most aggressive operating point, KeepAD achieves a $7.9\times$ speedup over the strongest CLIP-based baseline.

arXiv

Oh Deer, How Should I Handle This? Seasonal Priors for Selective Wildlife Annotation and Classification

Hugo Markoff, Christoph Praschl, Anton Hjalte Jørgensen, Christian Emil Mogensen, Mathias Bech Skadhauge, Sara Beery, Michael Ørsted, David C. Schedl

Fine-grained wildlife classification in aerial imagery is limited not only by model performance, but also by unreliable labels: animals occupy few pixels, key visual cues vary seasonally, and modality-specific evidence can be ambiguous. We study adult-male identification in red deer ($\textit{Cervus elaphus}$), where the antler cycle defines predictable windows of reliable evidence for both annotation and prediction. Using 7,295 RGB-only, thermal-only, and matched RGB+thermal crop sets from low-altitude UAV surveys, labeled by three annotators, we show that seasonal structure links (I) annotation quality, (II) downstream classification, and (III) selective prediction. Matched RGB+thermal review resolves more samples than either single modality, recovering majority-male labels otherwise missed by RGB or thermal alone, in human-based as well as model-based classification. Months with high annotator abstention also show lower classifier confidence, and soft seasonal priors mainly benefit the season-limited thermal view. Uncertainty-band abstention further raises covered accuracy to 98.9%, though at reduced coverage and with deferral that falls disproportionately on males. Overall, a biologically grounded seasonal calendar predicts where annotation and prediction are unreliable, and can guide both annotation protocol design and modality weighting.

arXiv

Hermite Curves as Trajectory Priors for Vision-Language-Action Models

Qi Lv, Jianming Xing, Zhao Yang, Mingyuan Yao, Yinan Shi, Yawei Jueluo, Mike Zheng Shou, Xiang Deng

Despite recent progress in Vision-Language-Action (VLA) models for robotic manipulation, the action chunk remains a weakly structured interface. Existing work typically flatten each chunk into per-timestep controls, relying on implicit data learning that manifests as jagged motion and boundary discontinuities during physical execution. To address these limitations, we introduce Hermite trajectory priors, parameterizing the chunk trajectory as a piecewise cubic Hermite curve defined by endpoint positions and velocities to explicitly enforce smoothness and continuity. We instantiate this fixed operator across discrete autoregressive and continuous generative paradigms via three variants: (1) Hermite Tokens, which predict quantized boundary variables autoregressively; (2) Hermite Scaffold, which decomposes clean actions into a base scaffold and residuals; and (3) Hermite Regularization, which applies the prior strictly as an auxiliary training objective. Across simulation benchmarks and real-robot platforms, Hermite Regularization achieves superior performance among these three variants, improving π0.5 baseline success rates from 95.9% to 98.7% on LIBERO, 85.7% to 90.9% on LIBERO-plus, and 63.4% to 90.0% across four real-robot tasks without additional inference overhead. Trajectory analyses reveal that explicitly structuring trajectory priors serves most effectively as a learning inductive bias rather than a runtime constraint.

arXiv

Now You Have My Healthy Attention: A U-DiT for Brain-MRI Inpainting

Danilo Danese, Angela Lombardi, Tommaso Di Noia

The ASNR-MICCAI BraTS Local Synthesis (Inpainting) task asks for the anatomically plausible completion of healthy brain tissue within a masked region of a T1-weighted MRI, providing a tumor-free anatomical reference for downstream analysis. As the task is scored by distortion metrics (SSIM, PSNR, MSE), we build a deterministic regression model and focus on giving it inductive biases tailored to inpainting. Our network follows the U-DiT principle of performing self-attention on a downsampled token grid: a volumetric encoder-decoder imports long-range context through a downsampled global self-attention block with three-dimensional rotary position embeddings, while convolutions and skip connections preserve high-frequency detail. Two ideas drive our results. First, we constrain the attention so that occluded ("void") tokens attend only to known-healthy tokens of the same volume, with a learned bias toward each query's contralateral homologue, forcing the completion to be inferred from observed anatomy rather than from other unknown regions. Second, we add a contralateral-symmetry input that supplies the mirrored healthy hemisphere as a patient-specific prior; since the brain is approximately bilaterally symmetric and lesions are typically unilateral, this prior improves the distortion metrics at matched structural similarity. On the official BraTS-2026 validation leaderboard our submission reaches a mean healthy-region SSIM of $0.864$, PSNR of $24.7$\,dB and MSE of $4.6{\times}10^{-3}$ over $219$ cases. We further analyse the residual smoothness inherent to distortion-optimal regression and discuss its implications for anatomical realism.

arXiv

Embodied.cpp: A Portable Inference Runtime of Embodied AI Models on Heterogeneous Robots

Ling Xu, Borui Li, Hao Wu, Chuyu Han, Xiangyu Li, Mohan Hua, Shiqi Jiang, Ting Cao, Chuanyou Li, Sheng Zhong, Shuai Wang

Embodied AI models now span vision-language-action (VLA) models and world-action models (WAMs), but practical deployment remains fragmented across model-specific Python stacks, backend assumptions, and robot-side glue code, especially on heterogeneous edge devices. Existing inference runtimes are designed mainly for request-response serving and therefore do not satisfy the runtime contract of embodied deployment: multi-rate execution inside closed-loop control, latency-first batch-1 inference on heterogeneous hardware, and extensible embodied interfaces beyond fixed token I/O. We present Embodied$.$cpp, a portable C++ inference runtime for embodied models. Based on an architectural analysis of representative VLA models and WAMs, Embodied$.$cpp captures a shared execution path and organizes it into five layers: input adapters, sequence builders, backbone execution, head plugins, and deployment adapters. The runtime provides modular multi-rate execution, latency-first fused inference, and extensible operator and I/O support, enabling deployment across heterogeneous devices, robots, and simulators through one backend abstraction. We evaluate Embodied$.$cpp on three VLA and two WAM models, using normalized comparisons across Python and C++ quantization configurations. Overall, Embodied$.$cpp achieves 1.05x-2.70x inference speedups and 7\%-77\% lower VRAM relative to Python baselines, while maintaining near-baseline success for most configurations. These results show that Embodied$.$cpp improves deployment efficiency while preserving high control quality across diverse embodied model architectures. Project Link: https://github.com/SEU-PAISys/Embodied.cpp

arXiv

From Activation to Specificity: Automating Counterfactual Testing of Visual Representations in the Human Brain

Yuval Golbari, navve wasserman, Matias Cosarinsky, Roman Beliy, Aude Oliva, Antonio Torralba, michal Irani, Tamar Rott Shaham

Identifying which brain regions represent a visual concept in the human brain is a central challenge in neuroscience. Existing approaches have localized coarse functional regions (e.g., faces, places) through activation maximization, identifying regions that activate strongly for a target concept relative to other concepts. Yet strong activation alone does not establish that a region represents the concept itself, as responses may instead be driven by correlated visual or semantic cues. We introduce BrainTRACE (Testing Representations through Counterfactual Evidence), an automated framework that combines generative and brain models to synthesize controlled stimuli and validate neural representations through targeted counterfactual-specificity testing. Given a query specifying a concept of interest, our framework constructs targeted stimulus sets comprising concept images, counterfactual edits that remove the target concept while preserving other image content, and images with candidate correlated distractors. It then uses an image-to-fMRI encoding model to predict brain responses and searches for representations that respond specifically to the target concept over correlated alternatives. BrainTRACE returns validated candidate representations and proposes follow-up fMRI experiments to further test or extend its discoveries. Our approach successfully recovers known functional localizations and identifies new candidate representations across dozens of concepts, validated on both predicted and measured fMRI data. Critically, we show that without counterfactual evaluation, a large fraction of localizations would be false positives, confirming that activation alone is insufficient evidence of representation.

arXiv

HEART: Exploiting Head Heterogeneity in Sparse Attention for Video Diffusion

Xuzhe Zheng, Yuexiao Ma, Jing Xu, Xiawu Zheng, Rongrong Ji, Fei Chao

Sparse attention accelerates video diffusion by allowing each attention head to focus on only a small subset of interactions. Existing methods already construct head-specific sparse patterns conditioned on the input. However, we find that these heads also differ in two less obvious but practically important ways. First, for some heads, sparse attention masks remain stable across many denoising steps, whereas others change rapidly. Second, heads differ substantially in their sensitivity to sparsification: applying the same threshold can induce markedly different errors in the final denoising velocity. Ignoring these differences leads to redundant mask prediction and suboptimal threshold calibration across heads under a global sparsity budget. We present HEART, short for Heterogeneity-Exploiting Adaptive Refresh and Thresholding, a training-free framework that exploits both forms of head heterogeneity. First, Temporal Mask Reuse (TMR) uses a lightweight per-head query-key drift signal to determine whether a cached sparse mask remains reliable across denoising steps, refreshing it only when the drift exceeds a prescribed threshold. Second, Error-guided Budgeted Calibration (EBC) evaluates candidate thresholds offline using a frequency-weighted denoising-velocity error, and assigns each head an appropriate threshold under a global sparsity budget. HEART requires no retraining, weight modification, or sparse-kernel changes, and can be integrated directly into existing sparse-attention pipelines. Across Wan2.1-1.3B, Wan2.1-14B, and HunyuanVideo-13B, HEART consistently pushes the quality--efficiency frontier of advanced sparse attention methods such as XAttention and SVG2 outward.

arXiv

UniPCB: A Generation-Assisted Vision-Based Measurement Framework for PCB Defect Inspection

Huan Zhang, Lianghong Tan, Yichu Xu, Zishan Su, Jiangzhong Cao, Huanqi Wu, Linwei Zhu, Xu Zhang

Automated optical inspection (AOI) is a vision-based instrumentation process that detects and localizes physical defects on printed circuit boards (PCBs) from optically acquired images. Its reliability is often limited by scarce and class-imbalanced defect observations and by insufficient representation of small, low-contrast defects against dense circuit patterns. We propose UniPCB, a generation-assisted PCB inspection framework that combines controlled defect synthesis with task-specific detection. The generation branch derives heterogeneous edge, relative pseudo-depth, and text conditions from PCB images. A ScaleEncoder embeds the conditions at four U-Net resolutions, while a Condition Modulation block performs FiLM-style spatially adaptive fusion. The detection branch introduces an Inverted Residual Shift Attention block for joint global-local modeling and a Cross-level Complementary Fusion block for selective hierarchical feature integration. The generated images augment the detector training data and improve defect coverage without changing the online inspection path. Experiments using the corrected annotations of DsPCBSD+ yield an mAP@0.5 of 98.0% and an mAP@0.5:0.95 of 61.8%. Under the same generation-augmented training setting, these values exceed those of the RT-DETR baseline by 2.1 and 1.7 percentage points, respectively. The generation branch obtains an FID of 129.61 and an SSIM of 0.619. These results indicate the potential of structured synthetic data and PCB-oriented feature modeling for improving vision-based defect inspection. Code is available at https://github.com/House-yuyu/UniPCB.

arXiv

Uncertainty in a Single Pass: A Closed-Form Identity for One-Step Flow Matching

Jiarui Xing, Song Wang, Jian Wang

Flow matching provides a highly effective framework for generative modeling, yet estimating the uncertainty of its generated samples remains a fundamental challenge. Existing methods rely on auxiliary variance heads, model ensembles, or iterative covariance propagation, which invariably introduce accumulated errors and structural interference along the noise-to-image trajectory. We resolve these limitations by adapting Tweedie's formula to the linear flow matching interpolant, yielding an exact, closed-form identity for the posterior covariance. Because this identity relies exclusively on the divergence of the learned velocity field, it can be evaluated post hoc on pretrained models without any architectural modifications. This analytical formulation proves exceptionally powerful for single-step generative frameworks. By computing the end-to-end posterior covariance in one forward pass, our approach completely bypasses the compounding errors inherent to sequential integration. This mechanism fundamentally enhances robustness, providing highly accurate and stable uncertainty estimates while preventing the numerical degradation typical of multi-step approximations. We validate our framework across benchmark datasets, including CIFAR-10 and real brain MRI scans. The resulting divergence-based uncertainty maps are highly interpretable and tightly correlated with empirical reconstruction errors. Crucially, in medical imaging, this precise localization of structural ambiguity provides essential decision support for high-stakes clinical tasks such as tumor boundary delineation and neurosurgical planning.

arXiv

Compatibility of Face Embeddings Across Deep Neural Networks

Fizza Rubab, Yiying Tong, Arun Ross

Automated face recognition has made rapid strides over the past decade due to the unprecedented rise of deep neural network (DNN) models that can be trained for domain-specific tasks. At the same time, large foundation models that are pretrained on broad vision or vision-language tasks have shown impressive generalization across diverse domains, including biometrics. This raises an important question: Do different DNN models---both domain-specific and foundation models---encode facial identity in similar ways, despite being trained on different datasets, loss functions, and architectures? In this regard, we directly analyze the geometric structure of embedding spaces imputed by different DNN models. Treating embeddings of face images as point clouds, we study whether simple affine transformations can align face representations of one model with another. Our findings reveal substantial cross-model compatibility: low-capacity linear mappings substantially improve cross-model face recognition over unaligned baselines for both identification and verification, including across foundation models never trained for face recognition. Alignment patterns generalize across datasets and vary systematically across model families, indicating representational convergence in facial identity encoding. These findings reframe independently trained templates as transferable rather than revocable, with implications for interoperability, ensemble design, and biometric template security.

arXiv

Let Geometry GUIDE: Layer-wise Unrolling of Geometric Priors in Multimodal LLMs

Chongyu Wang, Ting Huang, Chunyu Sun, Xinyu Ning, Di Wang, Hao Tang

Multimodal Large Language Models (MLLMs) have achieved remarkable progress in 2D visual tasks but still struggle to understand physical space in real-world visual streams. Recently, feed-forward geometric foundation models that implicitly extract geometric priors from visual inputs have offered a promising direction. However, many existing geometry-aware MLLMs extract features from a single deep encoder layer and perform one-shot fusion at the language-model input, potentially overlooking complementary information across geometric encoder layers, limiting access to fine-grained geometric cues, and hindering progressive cross-modal adaptation. To address these limitations, we propose GUIDE (Geometric Unrolling Inside MLLM Early-layers), a progressive framework for integrating geometric priors. GUIDE samples and aligns multi-level features from the geometric encoder and sequentially injects them into the early layers of the MLLM. This design enables the MLLM to continuously access multi-granularity geometric cues and progressively integrate them with visual representations during layer-wise semantic abstraction. GUIDE further introduces a dual context-aware gating mechanism at both the token and layer levels, selectively regulating geometric information to reduce redundant injection and interference with pretrained representations. Extensive experiments on spatial reasoning and 3D scene understanding benchmarks, including VSI-Bench, ScanRefer, and Scan2Cap, validate the effectiveness and cross-task applicability of GUIDE. Our 5B and 9B models achieve average scores of 71.5 and 72.1 on VSI-Bench, respectively.

arXiv

Dense Point-to-Mask Optimization with Reinforced Point Selection for Crowd Instance Segmentation

Hongru Chen, Jiyang Huang, Jia Wan, Antoni B. Chan

Crowd instance segmentation is a crucial task with a wide range of applications, including surveillance and transportation. Currently, point labels are common in crowd datasets, while region labels (e.g., boxes) are rare and inaccurate. The masks obtained through segmentation help to improve the accuracy of region labels and resolve the correspondence between individual location coordinates and crowd density maps. However, directly applying currently popular large foundation models such as SAM does not yield ideal results in dense crowds. To this end, we first propose Dense Point-to-Mask Optimization (DPMO), which integrates SAM with the Nearest Neighbor Exclusive Circle (NNEC) constraint to generate dense instance segmentation from point annotations. With DPMO and manual correction, we obtain mask annotations from the existing point annotations for traditional crowd datasets. Then, to predict instance segmentation in dense crowds, we propose a Reinforced Point Selection (RPS) framework trained with Group Relative Policy Optimization (GRPO), which selects the best predicted point from a sampling of the initial point prediction. Through extensive experiments, we achieve state-of-the-art crowd instance segmentation performance on ShanghaiTech, UCF-QNRF, JHU-CROWD++, and NWPU-Crowd datasets. Furthermore, we design new loss functions supervised by masks that boost counting performance across different models, demonstrating the significant role of mask annotations in enhancing counting accuracy.

arXiv

A Comparative Study in Surgical AI: Potential and Limitations of Data, Compute, and Scaling

Kirill Skobelev, Eric Fithian, Yegor Baranovski, Jack Cook, Sandeep Angara, Shauna Otto, Zhuang-Fang Yi, John Zhu, Neeraj Mainkar, Margaux Masson-Forsythe, Daniel A. Donoho, X. Y. Han

Recent Artificial Intelligence (AI) models have matched or exceeded human experts in several benchmarks of biomedical task performance, but surgical benchmarks in particular are often missing from prominent medical benchmark suites. Since surgery requires integrating disparate tasks, generally-capable AI models could be particularly attractive as a collaborative tool if performance could be improved. On the one hand, the canonical approach of scaling architecture size and training data is attractive, especially since there are millions of hours of surgical video data generated per year. On the other hand, preparing surgical data for AI training requires significantly higher levels of professional expertise, and training on that data requires expensive computational resources. These trade-offs paint an uncertain picture of whether and to-what-extent modern AI could aid surgical practice. In this paper, we explore this question through a case study of surgical tool detection using state-of-the-art AI methods available in 2026. We demonstrate that even with multi-billion parameter models and extensive training, current Vision Language Models fall short in the seemingly simple task of tool detection in neurosurgery. Additionally, we show scaling experiments indicating that increasing model size and training time only leads to diminishing improvements in relevant performance metrics. Thus, our experiments suggest that current models could still face significant obstacles in surgical use cases. Moreover, some obstacles cannot be simply ``scaled away'' with additional compute and persist across diverse model architectures, raising the question of whether data and label availability are the only limiting factors. We discuss the main contributors to these constraints and advance potential solutions.

arXiv

Where and How to Prune: An Empirical Study of Visual Token Pruning for GUI Agent Navigation

Daiqiang Li, Zihao Pan, Zeyu Zhang, Xuyang Liu, Shijia Xu, Ronghao Chen, Huacan Wang, Honggang Chen, Linfeng Zhang, Zhangquan Chen, Haiyun Jiang

In recent years, GUI agents have demonstrated strong potential in navigation tasks. However, preserving complete historical screenshots introduces substantial computational overhead. This paper investigates how token pruning, a plug-and-play inference acceleration technique, can be effectively applied to GUI agent navigation scenarios. Firstly, we address the question of where pruning should occur. We identify a system-level redundancy overlooked by existing methods: as the same screenshot is repeatedly fed into the model across different steps, its ViT encoding is redundantly recomputed each time. We show that its ViT-encoded embeddings can be fully cached and reused across steps, substantially reducing FLOPs while preserving model performance. This finding suggests that inference acceleration efforts should focus on the subsequent Large Language Model (LLM). Building on this, we further address the question of how to prune within the LLM, and distill two key insights: (i) from a semantic perspective, the token budget should be balanced between foreground and background regions; (ii) from a spatial perspective, the spatial uniformity of retained tokens should be maintained to preserve the model's global spatial perception. These findings provide practical guidance for the design of inference acceleration and token pruning for GUI agent navigation.

arXiv

PolypSteer: Counterfactual Endoscopic Synthesis via Training-Free Activation Steering

Trong-Thang Pham, Loc Nguyen, Anh Nguyen, Hien V. Nguyen, Ngan Le

Generative diffusion models are increasingly used for medical imaging data augmentation, but text prompting cannot produce causal training data. Re-prompting rerolls the entire generation trajectory, altering anatomy, texture, and background. Inversion-based editing methods introduce reconstruction error that causes structural drift. We propose PolypSteer, a training-free activation-steering framework for endoscopic synthesis. PolypSteer identifies a pathology vector for each contrastive prompt pair in the cross-attention layers of a diffusion transformer. At inference time, it steers image activations along this vector, generating counterfactual pairs from scratch where the only difference is the steered concept. All other structure is preserved by construction. We evaluate PolypSteer across three experiments on Kvasir v3 and HyperKvasir. On counterfactual generation across three clinical concept pairs, PolypSteer achieves flip rates of 0.800, 0.925, and 0.950, outperforming the best inversion-based baseline in both concept flip rate and structural preservation. On dye disentanglement, PolypSteer achieves 75% dye removal against 20% (PnP) and 10% (h-Edit). On downstream polyp detection, augmenting with PolypSteer counterfactual pairs achieves ViT AUC of 0.9755 versus 0.9083 for quantity-matched re-prompting, confirming that counterfactual structure drives the gain. Code is at link https://github.com/UARK-AICV/PolypSteer

继续浏览

更多 AI 顶会论文

30+ 个顶会，按会议、年份和主题浏览。

浏览全部会议

进入任一会议和年份，查看完整论文列表。

全部会议 →

搜索论文

按标题、作者或关键词，在 21 万+ 篇论文里搜索。

搜索论文 →