cs.CV（2026-09-18）

📊 共 30 篇论文 | 🔗 9 篇有代码

🎯 兴趣领域导航

支柱三：空间感知与语义 (Perception & Semantics) (11 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (8 🔗4) 支柱九：具身大模型 (Embodied Foundation Models) (6 🔗1) 支柱一：机器人控制 (Robot Control) (4 🔗3) 支柱四：生成式动作 (Generative Motion) (1)

🔬 支柱三：空间感知与语义 (Perception & Semantics) (11 篇)

#题目一句话要点标签🔗⭐
1 VoxelTTO: Voxel-Aligned Feed-Forward 3D Gaussian Splatting with Test-Time Optimization 提出VoxelTTO以解决3D高斯点云重建中的对齐问题 3D gaussian splatting 3DGS gaussian splatting
2 Cube-Splat: High-Fidelity 360° Gaussian Splatting SLAM via Cubemap Factorization and Adjoint-Consistent Optimization 提出Cube-Splat以解决全景图像下的高保真SLAM问题 visual SLAM 3D gaussian splatting 3DGS ✅
3 4DGS-Fixer: Generative Sparse-View 4D Gaussian Splatting with Iterative Refinement Guided by Video Diffusion Priors 提出4DGS-Fixer以解决动态场景合成中的稀疏视图问题 gaussian splatting splatting
4 2D GauSS-MI: Efficient Active Scene Reconstruction with Balanced Visual and Geometric Quality 提出2D GauSS-MI以解决高效场景重建问题 gaussian splatting splatting scene reconstruction
5 SFVO: Decoupled Confidence-Guided Stereo-Flow Visual Odometry with Bidirectional PnP 提出SFVO以解决立体视觉里程计中的几何约束问题 visual odometry optical flow
6 Info3R: Information-Adaptive Test-Time Training for 3D Reconstruction 提出Info3R以解决长图像序列中的3D重建问题 depth estimation 3D reconstruction
7 Think Locally, Refine Globally for Memory-Efficient 3D Reconstruction 提出LoG-VGGT以解决长序列3D重建中的内存效率问题 3D reconstruction VGGT
8 VGGT-CAD: Reconstructing Parametric CAD 3D Model with Geometric Grounding 提出VGGT-CAD以解决参数化CAD重建中的几何约束问题 VGGT
9 XCalib Depth-Guided Geometric Optimization for Dense Thermal-Visible Video Registration 提出XCalib以解决热成像与可见光视频注册问题 metric depth multimodal
10 Object Detection Benchmarks are Incomplete: The Role of Label Errors and Annotation Uncertainty 提出高召回率的不确定性感知标注方法以解决目标检测基准不完整问题 open-vocabulary open vocabulary
11 VeriFuse: Bounded Vision-Language Arbitration and Reason-Guided Refinement for Cooperative 3D Perception 提出VeriFuse以解决车辆与基础设施的协同3D感知问题 scene understanding

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#题目一句话要点标签🔗⭐
12 DRT: Dense Reasoning Trace for Efficient and Grounded Multimodal Reasoning 提出DRT以解决多模态推理中的语言冗余问题 reinforcement learning large language model multimodal ✅
13 ZYT-World: A Real-Time Controllable World Model for Closed-Loop Autonomous-Driving Simulation 提出ZYT-World以解决闭环自主驾驶仿真中的三大挑战 world model world models distillation
14 Adaptive World Memory 3D Foundation Model for Scalable 3D Mapping, Localization, and Rendering 提出自适应世界记忆3D基础模型以解决可扩展3D映射与定位问题 world model world models foundation model ✅
15 JEPA Guided Diffusion: Predictive Vision-Language Conditioning for Generative Traffic Forecasting 提出JEPA引导扩散框架以解决交通预测问题 JEPA representation learning scene understanding
16 GestureFAR: Streaming Co-Speech Gesture Generation with Flow Autoregression 提出GestureFAR以解决流媒体共语手势生成问题 flow matching distillation motion latent ✅
17 PointLAM: Local Attentive Mamba for Efficient Point-based 3D Object Detection 提出PointLAM以解决点云3D目标检测的效率与精度问题 Mamba ✅
18 S3VD: Semantic-Guidance Spatio-Temporal Scanning for Video Deraining 提出S3VD框架以解决视频降雨问题 Mamba SSM state space model
19 RobotEQ-Video: A Video-Centric Benchmark for Social Proactive Intelligence with World-State Taxonomy 提出RobotEQ-Video以解决社交主动智能研究中的视频分析不足问题 world model world models

🔬 支柱九：具身大模型 (Embodied Foundation Models) (6 篇)

#题目一句话要点标签🔗⭐
20 PRIME: Perception Feedback with Situational Memory Embeddings in VLA Models 提出PRIME以解决VLA模型感知反馈不足问题 vision-language-action VLA
21 AgentVidBench: A Multi-Hop Video Question Answering Benchmark for Evaluating MLLM Agents 提出AgentVidBench以解决多跳视频问答评估问题 large language model multimodal ✅
22 SignGPT: Toward LLM-Mediated Sign Language Interaction through Gloss-Free Translation and Generation 提出SignGPT以解决手语交互中的模型切换问题 large language model
23 From Retrieval to Recognition:How Vision--Language Models Become OCR Specialists 提出新方法使视觉语言模型具备OCR专长 multimodal
24 Refine Then Fusion: Training-Free 3D Point Cloud Adaptation with Priority Refinement and Multi-Modal Knowledge Fusion 提出RTF框架以解决少样本3D识别中的多模态融合问题 foundation model
25 Hiding in Plain Sight: A Diffusion-based Mitigation of Geolocation Privacy Leakage in Vision-Language Models 提出基于扩散模型的框架以解决多模态模型地理位置隐私泄露问题 multimodal

🔬 支柱一：机器人控制 (Robot Control) (4 篇)

#题目一句话要点标签🔗⭐
26 ME-Dex 1.0: Bringing Heterogeneous Tactile Sensing into World Action Modeling 提出ME-Dex 1.0以解决异构触觉感知在世界动作建模中的不足 manipulation dexterous hand flow matching
27 A Scene Language Model for Open-Vocabulary Scene Mapping 提出SceneLM以解决开放词汇3D场景映射问题 quadruped open-vocabulary open vocabulary ✅
28 Hand-Aware Transition Modeling for Bimanual Procedural Anomaly Detection 提出HACT模型以解决双手操作中的程序异常检测问题 bi-manual ✅
29 Adaptive Color Grading 提出自适应色彩分级工具以优化图像处理 manipulation ✅

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
30 Geometry-Aware Diffusion Guidance via Curvature-Adaptive Tubular Correction 提出曲率自适应管状校正以优化扩散引导 classifier-free guidance

⬅️ 返回 cs.CV 首页 · 🏠 返回主页