cs.CV（2026-10-09）

📊 共 47 篇论文 | 🔗 8 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (15 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (13 🔗3) 支柱三：空间感知与语义 (Perception & Semantics) (10 🔗3) 支柱四：生成式动作 (Generative Motion) (3) 支柱一：机器人控制 (Robot Control) (3) 支柱八：物理动画 (Physics-based Animation) (2) 支柱六：视频提取与匹配 (Video Extraction) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (15 篇)

#题目一句话要点标签🔗⭐
1 Measurable Semantic 3D Gaussian Splatting with Joint Region and Depth Supervision for Interactive Building Inspection 提出可测量的语义3D高斯点云以解决建筑检查中的交互性问题 MAE 3D gaussian splatting 3DGS
2 Knowing Whether, Not Just Where: Anchor-Probing Distillation for Open-Set Remote Sensing Visual Grounding 提出Anchor-Probing Distillation解决开放集遥感视觉定位问题 teacher-student distillation visual grounding
3 OWLET: On-Policy Weighted Learning via Evidence Transfer for Video Large Language Model Post-Training 提出OWLET以解决视频大语言模型后训练中的视觉证据利用问题 reinforcement learning distillation spatiotemporal
4 Shortcut Learning in Geospatial Foundation Model Embeddings for Cross-Disaster Damage Assessment 提出地理基础模型嵌入以解决跨灾害损伤评估中的快捷学习问题 contrastive learning foundation model ✅
5 A Lightweight Velocity Correction Framework for Improving Instruction Following in Few-Step Text-to-Image Generation 提出轻量级速度修正框架以改善少步文本到图像生成的指令跟随能力 distillation instruction following
6 GLIMPSE: Global-Local Indexing with Multi-Precision Spectral Encoding for Explorable Video World Models 提出GLIMPSE以解决视频世界模型中的位置精度问题 world model world models
7 A4-MR: Abnormality-aware Anatomical Alignment for MR 提出A4-MR以解决MRI图像解剖定位与病理表示不足的问题 representation learning foundation model multimodal
8 Selective Spatio-Temporal Fruit Inspector (S$^2$FI): A Mamba-based Framework for Rotational Fruit Quality Assessment 提出S$^2$FI框架以解决果蔬分拣效率低下问题 Mamba spatiotemporal
9 The Predictability Bias of Joint-Embedding Predictive Learning in Medical Imaging 提出TAP-JEPA以解决医学影像中的可预测性偏差问题 JEPA Joint-Embedding Predictive Architecture joint-embedding predictive architecture
10 X-AVDD: Cross-Attentive Audio-Visual Dataset Distillation 提出X-AVDD以解决音视频数据蒸馏中的内存和语义对应问题 distillation
11 QCLR: A Plug-in Robustification Module via Quantization for Contrastive Learning in Self-Supervised Skeleton-Based Human Action Recognition 提出QCLR以增强自监督骨架人类动作识别的鲁棒性 contrastive learning
12 ConClick: Element-Wise Contrastive Learning Improves GUI Grounding 提出ConClick框架以提升GUI元素定位精度 contrastive learning ✅
13 CV-OPD: Contrastive Vision On-policy Distillation 提出CV-OPD以解决视觉语言模型中的蒸馏短路问题 distillation
14 Redirecting Teacher Supervision: Temporal-Response-Guided Latent Distillation for Video-Language Models 提出时间响应引导的潜在蒸馏方法以提升视频语言模型性能 distillation
15 CoVA: Composing VAE Representations for Generative Learning 提出CoVA以优化生成学习中的VAE表示 flow matching classifier-free guidance

🔬 支柱九：具身大模型 (Embodied Foundation Models) (13 篇)

#题目一句话要点标签🔗⭐
16 Jag: Direct Box Prediction for Efficient Visual Grounding 提出Jag以解决视觉定位中的效率与准确性问题 multimodal visual grounding ✅
17 The Evidence Anchoring Boundary: Where Visual Interventions Matter and When They Help in Multimodal LLMs 提出证据锚定边界以优化多模态大语言模型的视觉干预 large language model multimodal
18 Can Multimodal Decision Models Classify Breast Ultrasound Images? Benchmarking CLEF and CLEF-Flash Across Prompts and Datasets 基于多模态决策模型的乳腺超声图像分类研究 multimodal
19 S-Space: Exploring Spatial Workspace in Multimodal Language Models 提出S-Space以揭示多模态语言模型的空间推理机制 multimodal
20 Dataset-Specific Encoding of Image Content Underlies Cross-Dataset Failure in Brain MRI Foundation Models 提出新方法解决脑MRI基础模型跨数据集失效问题 foundation model ✅
21 MotionCoach: Constructing Multimodal Guidance for Video Professional Motion Editing 提出MotionCoach以解决专业运动编辑中的多模态指导问题 multimodal
22 Video Codec-Guided Training-Free Token Compression for Video Large Language Models 提出视频编码引导的无训练令牌压缩方法以提升视频语言模型性能 large language model
23 Stride: Scale-Aware Multi-Agent Evolutionary Refinement for Infographic Generation 提出Stride框架以解决信息图生成中的多代理进化优化问题 multimodal visual grounding
24 SheafPrune: Sheaf-Guided Graph Propagation for Visual Token Pruning 提出SheafPrune以解决视觉标记选择中的效率问题 large language model multimodal ✅
25 From Glance to Focus: Mitigating Recognition Laziness in MLLMs for Fine-Grained Recognition 提出LIVE框架以解决MLLMs在细粒度识别中的懒惰问题 large language model multimodal
26 SpatialASM: Internalizing Active Mentalization for Spatial Reasoning 提出SpatialASM以解决大型视觉语言模型的空间推理问题 chain-of-thought
27 GuidedNav: Shaping Vision-Language Navigation Representations through Pre-action Attention and Spatial Guidance 提出GuidedNav以解决视觉语言导航中的决策指导问题 VLN
28 Robust Vision-Language Navigation with Awareness of Instruction-Execution Progress 提出VLN-AIEP框架以解决多步指令导航中的执行进度干扰问题 VLN

🔬 支柱三：空间感知与语义 (Perception & Semantics) (10 篇)

#题目一句话要点标签🔗⭐
29 NovaSplat: Model-Free 6D Object Pose Estimation via 3D Gaussian Splatting Onboarding and Refinement 提出NovaSplat以解决无CAD模型的6D物体姿态估计问题 3D gaussian splatting 3DGS gaussian splatting ✅
30 Make It Flow: Animating Water in 3D Gaussian Splatting Scenes 提出水流动画方法以解决静态3D场景动态表现不足问题 3D gaussian splatting 3DGS gaussian splatting
31 Gaussian Splatting with Reflections Guided by What is Seen 提出基于2D高斯点云的新框架以解决近场反射问题 3D gaussian splatting 3DGS gaussian splatting
32 SplatQ: Interaction-Aware Pruning of 3D Gaussian Splatting via Quadratic Selection 提出SplatQ以解决3D高斯点云剪枝中的区域预算控制问题 3D gaussian splatting 3DGS gaussian splatting
33 CompactSplat: Spatially Adaptive Gaussian Distribution for Feedforward Scene Reconstruction 提出CompactSplat以解决稀疏视图下3D场景重建问题 3D gaussian splatting 3DGS gaussian splatting
34 Efficient Confidence-Aware Densification for 3DGS 提出基于置信度的高效三维稠密化方法以解决重建不足问题 3D gaussian splatting 3DGS gaussian splatting
35 PhysForge: Material-Controllable 3D Reconstruction with Separable Components 提出PhysForge以解决物理属性控制不足的问题 3D reconstruction ✅
36 MeshBind: Point-to-Mesh Binding for Non-Rigid Scene Flow Automatic Annotation 提出MeshBind以解决非刚性场景流自动标注问题 scene flow
37 Volume Rendering with Metric--Phase Fields for Thin Structures and Sign-Ambiguous Geometry 提出PhaseSurf以解决薄结构和符号模糊几何体的体积渲染问题 3DGS
38 AD-SRF: Attack-Defense Semantic Reinforcement for Multi-Modal Image Fusion 提出AD-SRF以解决多模态图像融合中的视觉质量与语义属性平衡问题 semantic map ✅

🔬 支柱四：生成式动作 (Generative Motion) (3 篇)

#题目一句话要点标签🔗⭐
39 Sonic-AMG: Streaming Human Motion Generation from Spatial Audio 提出Sonic-AMG以解决空间音频驱动的人体动作生成问题 motion generation motion latent human motion
40 ELMA: Benchmarking Anaphoric Compositions for Long-Term Text-to-Motion Generation 提出ELMA以解决长文本指令下的人类动作生成问题 text-to-motion motion synthesis motion generation
41 Towards Physically Grounded Articulated 3D Assets: A Multi-Agent Pipeline and Benchmark 提出Artisan系统以生成物理基础的关节3D资产 physically plausible

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
42 SimVTA: Simulation-Grounded Video-Trajectory Alignment for Driving World Action Models 提出SimVTA以解决驾驶世界行动模型的视觉监督问题 trajectory optimization reinforcement learning flow matching
43 What Visual Representations Do World Action Models Need? 提出R$^2$-WAM以优化世界行动模型的视觉表示 manipulation world action model world action models
44 LAPT-Scale: Scaling Embodied Policies with Latent Action Pre-Training on Human Videos 提出LAPT-Scale以解决机器人政策扩展问题 manipulation optical flow egocentric

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

#题目一句话要点标签🔗⭐
45 READ: Reading the Selective Scan for Faithful Video Attribution 提出READ以解决长序列视频建模的可解释性问题 spatiotemporal
46 TextVBench: Evaluating Text Rendering and Preservation in Video Generation 提出TextVBench以解决视频生成中文本渲染与保留问题 spatiotemporal

🔬 支柱六：视频提取与匹配 (Video Extraction) (1 篇)

#题目一句话要点标签🔗⭐
47 What Egocentric Video Embeddings Encode and How to Combine Them with Inertial Sensors for Human Activity Recognition 提出结合自我中心视频与惯性传感器的人类活动识别方法 egocentric

⬅️ 返回 cs.CV 首页 · 🏠 返回主页