cs.CV（2026-06-10）

📊 共 39 篇论文 | 🔗 10 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (15 🔗3) 支柱三：空间感知与语义 (Perception & Semantics) (9 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (9 🔗1) 支柱一：机器人控制 (Robot Control) (3 🔗1) 支柱四：生成式动作 (Generative Motion) (1) 支柱七：动作重定向 (Motion Retargeting) (1 🔗1) 支柱六：视频提取与匹配 (Video Extraction) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (15 篇)

#题目一句话要点标签🔗⭐
1 VLGA: Vision-Language-Geometry-Action Models for Autonomous Driving 提出VLGA模型以解决自动驾驶中的3D场景理解问题 vision-language-action VLA foundation model
2 From 2D Grids to 1D Tokens: Reforming Shared Representations for Multimodal Image Fusion 提出1D令牌接口以解决多模态图像融合中的全局一致性问题 multimodal ✅
3 Damage-TriageFormer: A Foundation-Model Framework for Typology-Based Building Damage Assessment from Mono-Temporal Imagery 提出Damage-TriageFormer以解决灾后建筑损伤评估问题 foundation model
4 Adapting Prithvi-EO for Fallow Detection for Food-Water Nexus: ViT-Adapter Necks and Parameter-Efficient Backbone tuning of Geospatial Foundation Model 提出ViT-Adapter以提高休耕地检测精度 foundation model
5 ParseFixer: An Agentic Framework for Document Parsing via Selective Multimodal Correction 提出ParseFixer以解决文档解析中的结构与内容恢复问题 multimodal ✅
6 MultiToP: Learning to Patch Visual Tokens to Mitigate Hallucinations in Video Large Multimodal Models 提出MultiToP以解决视频多模态模型中的幻觉问题 multimodal
7 Frozen Foundation-Model Embeddings Discard Small-Lesion Signal in Chest Radiography: Implications for Pre-Deployment Evaluation 提出冻结基础模型嵌入以解决胸部放射影像小病灶信号丢失问题 foundation model
8 Bridging the Modality Gap in Forensic Image Retrieval 提出统一检索框架以解决法医图像检索中的模态差异问题 large language model multimodal
9 Q-Fold: Query-Aware Focus-Context Spatio-Temporal Folding for Long Video Understanding 提出Q-Fold以解决长视频理解中的信息保留问题 large language model multimodal
10 Ouroboros-Spatial: Closing the Data-Model Loop for Spatial Reasoning 提出Ouroboros-Spatial以解决空间推理中的数据模型闭环问题 large language model multimodal
11 Findings of the MAGMaR 2026 Shared Task MAGMaR 2026共享任务成果展示与分析 multimodal
12 Task-Aware Structured Memory for Dynamic Multi-modal In-Context Learning 提出任务感知结构化记忆以解决多模态学习中的适应性问题 large language model
13 A Comprehensive Ecosystem for Open-Domain Customized Video Generation 提出PexelsCustom-1M数据集与CustoMDiT框架以解决开放域定制视频生成问题 multimodal
14 Adapting Vision-Language Models from Iconic to Inclusive for Multi-Label Recognition Without Labels 提出无标签多标签图像识别的框架以解决视觉语言模型的偏见问题 multimodal ✅
15 AVIS: Adaptive Test-Time Scaling for Vision-Language Models 提出AVIS以优化视觉语言模型的推理效率 chain-of-thought

🔬 支柱三：空间感知与语义 (Perception & Semantics) (9 篇)

#题目一句话要点标签🔗⭐
16 Scene-Adaptive Nonlinear Tone Curves for Pseudo Ground-Truth Generation in Low-Light 3D Gaussian Splatting 提出场景自适应非线性调色曲线以解决低光照3D重建问题 3D gaussian splatting 3DGS 3D reconstruction ✅
17 Wild3R: Feed-Forward 3D Gaussian Splatting from Unconstrained Sparse Photo Collection 提出Wild3R以解决稀疏照片集合中的3D高斯点云重建问题 3D gaussian splatting 3DGS gaussian splatting
18 Seeing What Matters: Perceptual Wrapper with Common Randomness for 3D Gaussian Splatting 提出感知包装器以解决3D高斯点云渲染中的纹理合成问题 3D gaussian splatting 3DGS gaussian splatting
19 DepthMaster: Unified Monocular Depth Estimation for Perspective and Panoramic Images 提出DepthMaster以解决单目深度估计在不同视角下的挑战 depth estimation monocular depth metric depth
20 LASA: A Weak Supervision Method for Open-Vocabulary Scene Sketch Semantic Segmentation 提出LASA方法以解决开放词汇场景草图语义分割问题 open-vocabulary open vocabulary
21 VL-DINO: Leveraging CLIP Vision-Language Knowledge for Open-Vocabulary Object Detectio 提出VL-DINO以解决开放词汇物体检测中的知识融合问题 open-vocabulary open vocabulary
22 Cross-Modal Benchmarking for Robotic Perception in Natural Environments 提出WildCross基准以解决自然环境下机器人感知问题 depth estimation metric depth foundation model ✅
23 Precision-Aware Illumination-Disentangled Vision Transformer for Spacecraft 6D Pose Estimation 提出精度感知的光照解耦视觉变换器以解决航天器6D姿态估计问题 6D pose estimation
24 Echoes of the Prior: A Computational Phenomenology of Forgetting 提出一种交互式安装以可视化遗忘的主观现象 3D reconstruction ✅

🔬 支柱二：RL算法与架构 (RL & Architecture) (9 篇)

#题目一句话要点标签🔗⭐
25 InternVideo3: Agentify Foundation Models with Multimodal Contextual Reasoning 提出InternVideo3以解决长视频多模态理解问题 reinforcement learning distillation foundation model
26 World Model Self-Distillation: Training World Models to Solve General Tasks 提出自蒸馏框架以解决视频生成模型的任务执行问题 reinforcement learning world model world models
27 Parameter-Efficient Adapter Tuning for Tabular-Image Multimodal Learning 提出TI-Adapter以解决多模态学习中的参数效率问题 predictive model multimodal
28 Slots, Transitions, Loops: Learning Composable World Models for ARC 提出Loop-OWM以解决ARC规则学习问题 world model world models
29 FreqKD: Frequency-Decoupled Cross-Modal Knowledge Distillation for Infrared Object Detection 提出FreqKD以解决红外目标检测中的知识蒸馏问题 distillation foundation model
30 Beyond Dark Knowledge: Mixup-Based Distillation for Reliable Predictions 提出基于Mixup的蒸馏方法以提升模型预测可靠性 distillation
31 AnchorEdit: Maintaining Temporal Consistency in Multi-turn Image Editing via Causal Memory 提出AnchorEdit以解决多轮图像编辑中的时间一致性问题 distillation instruction following
32 DynaTok: Token-Based 4D Reconstruction from Partial Point Clouds 提出DynaTok以解决部分点云序列的4D重建问题 flow matching spatiotemporal ✅
33 SpecLoR: Spectral Lookahead Rectification for Motion-Coherent Text-to-Video Generation 提出SpecLoR以解决文本到视频生成中的时空不一致问题 flow matching spatiotemporal

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
34 Making Foresight Actionable: Repurposing Representation Alignment in World Action Models 提出AGRA以解决世界动作模型中的表示不匹配问题 manipulation world model world models
35 TopoCap: Learning Topology-Agnostic Motion Priors for Monocular Video-to-Animation 提出TopoCap框架以解决单目视频到动画的运动捕捉问题 quadruped biped flow matching ✅
36 4DP-QA: Scalable QA for 4D Perception in Vision Language Models 提出4DP-QA以解决视觉语言模型在4D场景理解中的挑战 motion tracking scene understanding

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
37 TextHOI-3D: Text-to-3D Hand-Object Interaction via Discrete Multi-View Generation and Joint Mesh Optimization 提出TextHOI-3D以解决手-物体交互的3D生成问题 physically plausible penetration

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
38 SG2Loc: Sequential Visual Localization on 3D Scene Graphs 提出SG2Loc以解决复杂室内环境的视觉定位问题 spatial relationship ✅

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

#题目一句话要点标签🔗⭐
39 Reason, Then Re-reason: Cross-view Revisiting Improves Spatial Reasoning 提出跨视角重访方法以提升空间推理能力 egocentric ✅

⬅️ 返回 cs.CV 首页 · 🏠 返回主页