cs.CV（2026-10-27）

📊 共 50 篇论文 | 🔗 7 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (17 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (14 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (13 🔗3) 支柱一：机器人控制 (Robot Control) (4) 支柱八：物理动画 (Physics-based Animation) (2 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (17 篇)

#题目一句话要点标签🔗⭐
1 Vision-Language-Action or World-Action? Just Action Models 提出Just Action Model以解决控制策略知识继承问题 vision-language-action VLA ✅
2 Zooming without Zooming v2: From Seeing Details to Connecting Evidence for Generalizable Multimodal Perception 提出EvidenceConnect以解决多模态感知数据合成难题 embodied AI multimodal
3 ARISE: Authentic Reference-Informed Stability Enhancement for Continual Multimodal Forgery Detection 提出ARISE以解决持续多模态伪造检测中的稳定性问题 large language model multimodal
4 Zero-Shot Multimodal Event Spotting in Soccer Broadcasts 提出零-shot框架以解决足球赛事自动识别问题 multimodal
5 Zhengrong-Spatial-Aware 1.0: Advancing 3D Multimodal Intelligence 提出Zhengrong-Spatial-Aware 1.0以解决3D多模态智能问题 multimodal
6 Multibranch Ensemble Head in Vision Foundation Models for Efficient Cucumber Phenological Stage Classification 提出多分支集成头以解决小样本领域迁移问题 foundation model
7 Evaluation-Guided Demonstration Refinement for Vision-Language-Action Policies: The Winning Entry in the MetaFine Open Competition 提出基于评估指导的演示优化方法以提升视觉语言动作策略 vision-language-action
8 PRISM-BENCH: Diagnosing Grounded and Faithful Multimodal Reasoning Beyond Final-Answer Accuracy 提出PRISM-BENCH以解决多模态推理的准确性不足问题 multimodal
9 FrameForge: Harnessing Priors of Video Foundation Models for Unified Image Generation 提出FrameForge以解决统一条件图像生成问题 foundation model
10 Can Discrete Tokens Serve as Semantically Lossless Representations of Visual Signals? 提出因子化向量量化以提升离散视觉表示的语义保真度 foundation model multimodal ✅
11 As the Story Unfolds: Watching a Film and Identifying Characters as a Human Does 提出角色命名任务以解决电影角色识别问题 large language model multimodal
12 CodeCap-Bench: Can Code Serve as a Good Video Caption? 提出CodeCap-Bench以评估代码作为视频字幕的有效性 large language model multimodal
13 Latent Visual Contrastive Reasoning with Multi-Branch Exploration 提出分支感知的潜在视觉对比推理框架以解决复杂视觉推理问题 large language model multimodal
14 Decoupling Consistency and Conflict across MLLM Layers for Fake News Detection 提出多层次一致性与冲突解耦方法以解决假新闻检测问题 large language model multimodal
15 LearnPrunerV2: Learnable Progressive Token Pruning for Vision-Language Models 提出LearnPrunerV2以解决视觉语言模型的高计算开销问题 multimodal
16 Evaluating AlphaEarth Embeddings for Crop Classification: Separability, Temporal Dynamics, and Cross-Regional Transferability 提出AlphaEarth嵌入以提升作物分类的准确性与可迁移性 foundation model
17 GarmentSegAgent: Autonomous Garment Segmentation for Web Fashion Images via Textual Self-Refinement 提出GarmentSegAgent以解决网络时尚图像的自主服装分割问题 foundation model

🔬 支柱二：RL算法与架构 (RL & Architecture) (14 篇)

#题目一句话要点标签🔗⭐
18 VGIQA: Fine-Grained Image Quality Assessment with Visual Grounding Rewards 提出VGIQA以解决图像质量评估的局限性问题 reinforcement learning large language model visual grounding
19 VideoLVR: Latent Visual Reasoning over Grounded Video Evidence 提出VideoLVR以解决视频理解中的视觉遗忘问题 reinforcement learning large language model multimodal
20 A note on data-efficient finetuning of vision foundation models 提出数据高效微调方法以提升视觉基础模型性能 Mamba foundation model
21 GameVerse: A Minute-Scale Gameplay Dataset for Long-Horizon Interactive World Modeling 提出GameVerse以解决长时间互动世界建模数据稀缺问题 world model world models
22 PerceptRL: Making Visual Perception Rewardable in MLLMs from Answers to Interactions 提出PerceptRL以解决多模态大语言模型的视觉感知瓶颈问题 reward design large language model multimodal
23 Systematic Morphometric Brain Autoencoding for Representation Learning in Neurodegenerative Diseases 提出SYMBA框架以解决神经退行性疾病的MRI表征学习问题 representation learning foundation model
24 Hy-Nomas: Towards Free-form Interactive World Generation 提出Hy-Nomas以解决交互式世界生成中的历史信息保持问题 world model world models distillation
25 Nemotron-Labs-Diffusion-Uni: Elastic Self-Refining Unified Diffusion LLM with Cross-Paradigm On-Policy Distillation 提出Nemotron-Labs-Diffusion-Uni以解决视觉理解与生成任务的统一问题 distillation large language model
26 ALADDIN: Adversarial Latent Diffusion Distillation for One-Step Identity Swapping 提出ALADDIN以解决一阶段身份交换的效率与真实感问题 flow matching distillation
27 Uncovering Shortcut Amplification in Video RLVR through the Lens of Perceptual Uncertainty Collapse 提出Bifocal自蒸馏策略优化以解决视频RLVR中的快捷放大问题 reinforcement learning distillation multimodal
28 Learning to Look Beyond Screens: Transferable Visual Observation from GUI Grounding 提出LBS模型以解决GUI基础上可转移的视觉观察问题 reinforcement learning multimodal ✅
29 Multi-task Disentangling Fusion Mamba-CNN for MODIS Spectral-Temporal Unmixing with Endmember Variability Modeling 提出MDF-Mamba以解决MODIS光谱时间序列解混合问题 Mamba
30 Confident First, Reflective Second: Reinforcement Learning of Confidence-Preserving Reflection for Video Understanding 提出CoFire以解决视频理解中的信心侵蚀问题 reinforcement learning
31 Representing the Unseen: Rethinking Visual Imagination for Spatial Reasoning 提出ImagiRep与GEARS以优化空间推理中的视觉想象 world model world models

🔬 支柱三：空间感知与语义 (Perception & Semantics) (13 篇)

#题目一句话要点标签🔗⭐
32 ClusterSplat: Semantic Cluster Selection for 3D Visual Grounding in Gaussian Splatting 提出ClusterSplat以解决3D视觉定位中的目标选择问题 3D gaussian splatting 3DGS gaussian splatting ✅
33 GraftRay: Physically Based Mesh Insertion into Gaussian Splatting Scenes via Path Tracing 提出GraftRay以解决虚拟物体在高斯点云场景中插入的真实感问题 3D gaussian splatting 3DGS gaussian splatting
34 Perceiving and Predicting Dynamic Sports: Spatiotemporal Intelligence in Vision Models for Egocentric Tennis Video 提出时空智能模型以解决动态体育视频理解问题 metric depth egocentric spatiotemporal
35 GeoGSplat: Geometry-Grounded Non-Pixel-Aligned Gaussian Splatting 提出GeoGSplat以解决高分辨率视图合成中的冗余问题 3D reconstruction gaussian splatting splatting ✅
36 GCSR-OCC: Geometry Calibration and Semantic Routing for Open-Vocabulary Occupancy Prediction 提出GCSR-OCC以解决开放词汇占用预测中的几何校准与语义路由问题 open-vocabulary open vocabulary
37 Learning to Code the World: Harness Evolution for Compositional 3D Reconstruction 提出SceneEvolve以解决可编辑3D场景重建问题 3D reconstruction spatial relationship ✅
38 LingGS: Decoupling Feed-Forward Geometry Prediction from Camera Tracking for High-Fidelity Gaussian SLAM 提出LingGS以解决高保真高斯SLAM中的相机跟踪与几何预测问题 depth estimation gaussian splatting splatting
39 3D-SceneQ: Empowering 3D LLM With Query-Guided Adaptive Pruning and Multi-Modal Feature Enhancement 提出3D-SceneQ以解决3D场景理解中的噪声与信息缺失问题 scene understanding large language model multimodal
40 CTBakeGS: training-free baking of photorealistic CT volume renderings into 3D Gaussian splats 提出CTBakeGS以解决CT体积渲染训练时间长的问题 3DGS gaussian splatting splatting
41 GPA-VGGT: Geometry-Preserving Acceleration of VGGT 提出GPA-VGGT以解决VGGT在长序列中的效率问题 VGGT
42 Honeycomb-Surfels: Compact Hexagonal Primitives for Efficient Scene Rendering and Reconstruction 提出Hexels以解决资源受限环境下的3D场景渲染问题 3D reconstruction splatting
43 TriDifNet: A Unified Three-Task Diffusion Network for End-to-End RGB-T Perception 提出TriDifNet以解决RGB-T行人感知中的多任务优化问题 scene understanding
44 SAGE: Semantic-Augmented Geometric Experts for Point Cloud Registration 提出SAGE以解决低重叠点云配准中的几何歧义问题 scene reconstruction

🔬 支柱一：机器人控制 (Robot Control) (4 篇)

#题目一句话要点标签🔗⭐
45 SO-101 Bench: Accessible Spatial Manipulation 提出SO-101 Bench以解决语言条件下空间操控的基准问题 manipulation VLA language conditioned
46 LabWorld: Learning Interactive 4D Worlds for Scientific Experimentation 提出LabWorld框架以重建交互式4D实验室世界 real2sim scene reconstruction motion reconstruction
47 Adapting Video Models for Robot Control via Motion-guided Asymmetric Prediction 提出运动引导的不对称预测以提升机器人控制性能 manipulation world action model world action models
48 ScIntBench: Physical Interaction Plausibility in 3D Scene Generation 提出ScIntBench以解决3D场景生成中的交互可靠性问题 manipulation

🔬 支柱八：物理动画 (Physics-based Animation) (2 篇)

#题目一句话要点标签🔗⭐
49 Repurposing Video Diffusion Transformers for Cross-View Temporal Object Correspondence 提出XTrack以解决跨视角时间物体对应问题 spatiotemporal
50 LociArt: A Unified Paradigm for Free-Style Local Video Stylization 提出LociArt以解决自由风格局部视频风格化问题 spatiotemporal ✅

⬅️ 返回 cs.CV 首页 · 🏠 返回主页