cs.CV（2026-06-23）

📊 共 50 篇论文 | 🔗 15 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (19 🔗6) 支柱三：空间感知与语义 (Perception & Semantics) (13 🔗5) 支柱九：具身大模型 (Embodied Foundation Models) (11 🔗2) 支柱一：机器人控制 (Robot Control) (2) 支柱四：生成式动作 (Generative Motion) (2 🔗1) 支柱八：物理动画 (Physics-based Animation) (2) 支柱六：视频提取与匹配 (Video Extraction) (1 🔗1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (19 篇)

#题目一句话要点标签🔗⭐
1 Open-Vocabulary BEV Segmentation with 3D-Aware Geometric Constraints 提出开放词汇BEV分割以解决3D几何不一致问题 distillation gaussian splatting splatting ✅
2 PointVG-R: Internalizing Geometric Reasoning in MLLMs for Precise Pointing Localization via Visual Chain of Thought 提出PointVG-R以解决指向基础视觉定位问题 reinforcement learning spatial relationship large language model ✅
3 FLUX3D: High-Fidelity 3D Gaussian Generation with Diffusion-Aligned Sparse Representation 提出FLUX3D以解决高频视觉细节保留问题 representation learning 3D gaussian splatting 3DGS
4 Trimming the Long-Tail of Visual World Modeling Evaluation 提出Tailor-Bench以解决视觉世界建模中的长尾问题 world model world models affordance
5 Autonomous Video Generation with Counterfactual Controllability for Self-Evolving World Models 提出反事实可控性的视频生成方法以实现自演化世界模型 world model world models spatiotemporal
6 MambaRaw: Selective State Space Modeling for Efficient 4K Raw Image Reconstruction 提出MambaRaw以解决高效4K原始图像重建问题 Mamba SSM state space model ✅
7 OrbitForge: Text-to-3D Scene Generation via Reconstruction-Anchored Video Synthesis 提出OrbitForge以解决文本生成视频的3D场景一致性问题 distillation 3D gaussian splatting 3D reconstruction
8 Boosting Text-Driven Video Segmentation via Geometry-Aware Distillation 提出GeoLaV以解决文本驱动视频分割中的几何一致性问题 distillation geometric consistency spatiotemporal ✅
9 REDI-Match: Rotation-Equivariant Distillation for Efficient and Robust Dense Matching 提出REDI-Match以解决旋转不变性问题 distillation feature matching foundation model ✅
10 ViTexQA: A Multi-Frame Temporal Perception Dataset for Video Text Question Answering 提出ViTexQA以解决视频文本问答中的时序感知问题 reinforcement learning multimodal chain-of-thought
11 Jolia: Concept-Level Vision-Language Alignment for 3D CT Contrastive Learning 提出ConQuer方法以解决3D医学图像与文本对齐问题 contrastive learning foundation model
12 video-SALMONN-R$^3$: Learning to ReWatch, ReAsk, and ReAnswer for Efficient Video Understanding 提出video-SALMONN-R$^3$以解决视频理解中的信息缺失问题 reinforcement learning large language model chain-of-thought
13 Tri-Efficient Transfer Learning for Point Cloud Videos 提出PoinTriE以解决点云视频理解中的高成本和内存瓶颈问题 contrastive learning foundation model multimodal
14 An LMM for Precisely Grounding Elements in Documents 提出PreciseDoc以解决文档元素精确定位问题 reinforcement learning multimodal visual grounding
15 Heterogeneous Knowledge Distillation via Geometry Decoupling and Momentum-Aware Gradient Regulation 提出SPOFA框架以解决异构知识蒸馏中的训练不稳定问题 distillation
16 Lite Any Stereo V2: Faster and Stronger Efficient Zero-Shot Stereo Matching 提出Lite Any Stereo V2以解决高效零-shot立体匹配问题 distillation foundation model ✅
17 MATCH: Flow Matching for Multi-View Anomaly Detection 提出MATCH方法以解决多视角异常检测问题 flow matching
18 TuringViT: Making SOTA Vision Transformers Accessible to All 提出TuringViT以解决视觉变换器训练门槛问题 linear attention VLA
19 M^2C-EvDet: Multi-Domain Multi-Order Cross-Modal Knowledge Distillation for Event-based Object Detection 提出M^2C-EvDet以解决事件驱动物体检测中的跨模态知识蒸馏问题 distillation

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

#题目一句话要点标签🔗⭐
20 Geometry-Aware Style Transfer in 3D Gaussian Splatting 提出几何感知风格迁移框架以解决3D高斯点云风格化问题 3D gaussian splatting 3DGS gaussian splatting ✅
21 Compact Object-Level Representations with Open-Vocabulary Understanding for Indoor Visual Relocalization 提出OpenReLoc以解决室内视觉重定位问题 scene understanding open-vocabulary open vocabulary
22 UniRED: Unified RGB-D Video Frame Interpolation with Event Guidance 提出UniRED以解决RGB-D视频帧插值中的模糊与几何一致性问题 3D reconstruction scene understanding geometric consistency
23 Pocket-SLAM: Rendering-Area-Aware Pruning for Memory-Efficient 3DGS-SLAM 提出渲染区域感知修剪以解决3DGS-SLAM内存效率问题 3D gaussian splatting 3DGS gaussian splatting ✅
24 MM-TRELLIS: Point-Cloud Guided Multi-Modal 3D Vehicle Generation in Autonomous Driving 提出MM-TRELLIS以解决多模态传感器在3D车辆生成中的不足问题 3D gaussian splatting gaussian splatting splatting ✅
25 3DCarGen: Scalable 3D Car Generation via 3D-consistent Multi-view Synthesis 提出3DCarGen以解决3D车辆生成中的视角一致性问题 3D gaussian splatting gaussian splatting splatting
26 FLAT: Feedforward Latent Triangle Splatting for Geometrically Accurate Scene Generation 提出FLAT以解决现有3D场景生成的几何准确性问题 3DGS splatting
27 GeoT2V-Bench: Benchmarking 3D Consistency in Text-to-Video Models via 3D Reconstruction 提出GeoT2V-Bench以评估文本到视频模型的3D一致性问题 3D reconstruction VGGT
28 UniDrive: A Unified Vision-Language and Grounding Framework for Interpretable Risk Understanding in Autonomous Driving 提出UniDrive以解决自动驾驶中的风险理解问题 scene understanding large language model multimodal ✅
29 SignNet-1M: Large-Scale Multilingual Sign Language Video Dataset with Downstream Benchmarks 提出SignNet-1M以解决手语模型在真实场景中的鲁棒性问题 3D gaussian splatting 3DGS gaussian splatting
30 DDStereo: Efficient Dual Decoder Transformers for Stereo 3D Road Anomaly Detection 提出DDStereo以解决立体3D物体检测的实时性与开放集泛化问题 open-vocabulary open vocabulary
31 Spherical-to-ERP Epipolar Rectification for Single-Axis Disparity in 360 Stereo 提出球面到等距矩形投影以解决360度立体图像的视差估计问题 optical flow
32 AerialFusionMapNet: Online HD Map Construction with Aerial-Onboard BEV Fusion 提出AerialFusionMapNet以解决在线高清地图构建问题 scene understanding ✅

🔬 支柱九：具身大模型 (Embodied Foundation Models) (11 篇)

#题目一句话要点标签🔗⭐
33 Towards Fast and Effective Long Video Understanding of Multimodal Large Language Models via Adaptive Quasi-Gaussian Sampling 提出AdaQ以解决长视频理解中的高计算和内存占用问题 large language model multimodal ✅
34 Spectral Evolution-Guided Token Pruning in Multimodal Large Language Models 提出基于跨层谱演变的令牌剪枝以提升多模态大语言模型效率 large language model multimodal
35 Accelerating Multimodal Large Language Models with Prior-Corrected Token Reduction 提出Prior-Corrected Token Reduction以解决多模态大语言模型的效率问题 large language model multimodal
36 IV-CoT: Implicit Visual Chain-of-Thought for Structure-Aware Text-to-Image Generation 提出IV-CoT以解决结构感知文本到图像生成问题 large language model chain-of-thought
37 Latent Visual States for Efficient Multimodal Reasoning 提出EVA框架以解决多模态推理中的效率问题 multimodal
38 Flood Mapping from RGB imagery using a Vision Foundation Model 提出基于视觉基础模型的RGB图像洪水映射方法 foundation model
39 Agentic Collaborative Cognition for Zero-Shot 3D Understanding 提出协作多智能体框架以解决零-shot 3D理解问题 large language model multimodal
40 ActiveScope: Actively Seeking and Correcting Perception for MLLMs 提出ActiveScope以解决多模态大语言模型的感知问题 large language model multimodal ✅
41 FiCA: Feed-forward instant Gaussian Codec Avatars from a Single Portrait Image 提出FiCA以解决单张肖像生成逼真头像的问题 foundation model
42 Dual-Branch Cross-Projection Debiasing through Diffusion-based Disentanglement 提出双分支交叉投影去偏见方法以解决模型偏见问题 foundation model
43 VisChronos: Revolutionizing Image Captioning Through Real-Life Events 提出VisChronos以解决图像描述中的语义鸿沟问题 large language model

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
44 DriveStack-VLA: Render-Teacher Alignment for BEV-Based DeepStack Vision-Language-Action Model 提出DriveStack-VLA以解决VLA驾驶模型空间智能不足的问题 motion planning vision-language-action VLA
45 ForensicsTok: Forensics-Guided Tokenized Modeling for Image Tampering Localization 提出ForensicsTok以解决图像篡改定位问题 manipulation splatting large language model

🔬 支柱四：生成式动作 (Generative Motion) (2 篇)

#题目一句话要点标签🔗⭐
46 Social Structure Matters in 3D Human-Human Interaction Generation 提出社会结构建模方法以解决3D人际互动生成问题 text-to-motion motion generation physically plausible
47 PatternGSL: A Structured Specification Language for Template-Free and Simulation-Ready 3D Garments 提出PatternGSL以解决3D服装重建中的模板依赖问题 physically plausible ✅

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

#题目一句话要点标签🔗⭐
48 MotifGen: Spatiotemporal interpolation of misaligned satellite images via multi-source generative modeling, in an application to tropical cyclones 提出MotifGen以解决卫星图像时空插值问题 spatiotemporal
49 Training-free Cross-domain Few-shot Segmentation via Robust Semantic Representation and Matching 提出无训练跨域少样本分割框架以解决高计算成本问题 ASE foundation model

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

#题目一句话要点标签🔗⭐
50 EgoSAT: A Comprehensive Benchmark of Egocentric Streaming Interaction Understanding 提出EgoSAT以解决自我中心视频流交互理解问题 egocentric ✅

⬅️ 返回 cs.CV 首页 · 🏠 返回主页