cs.CV（2025-11-30）

📊 共 21 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱三：空间感知 (Perception & SLAM) (11 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱一：机器人控制 (Robot Control) (4 🔗2) 支柱五：交互与反应 (Interaction & Reaction) (2 🔗1)

🔬 支柱三：空间感知 (Perception & SLAM) (11 篇)

#题目一句话要点标签🔗⭐
1 Smol-GS: Compact Representations for Abstract 3D Gaussian Splatting Smol-GS：提出紧凑的抽象3D高斯溅射表示方法，实现高效场景压缩。 3D gaussian splatting 3DGS gaussian splatting
2 Feed-Forward 3D Gaussian Splatting Compression with Long-Context Modeling 提出基于长程上下文建模的前馈3D高斯溅射压缩方法，实现高压缩率。 3D gaussian splatting 3DGS gaussian splatting
3 PolarGS: Polarimetric Cues for Ambiguity-Free Gaussian Splatting with Accurate Geometry Recovery PolarGS：利用偏振信息实现无歧义高斯溅射和精确几何重建 3D gaussian splatting 3DGS gaussian splatting
4 EAG3R: Event-Augmented 3D Geometry Estimation for Dynamic and Extreme-Lighting Scenes EAG3R：事件相机增强的3D几何估计，解决动态和极端光照场景问题 SLAM depth estimation monocular depth
5 LAHNet: Local Attentive Hashing Network for Point Cloud Registration LAHNet：面向点云配准的局部注意力哈希网络，提升特征区分性。 point cloud interaction transformer
6 Dynamic-eDiTor: Training-Free Text-Driven 4D Scene Editing with Multimodal Diffusion Transformer Dynamic-eDiTor：基于多模态扩散Transformer的免训练文本驱动4D场景编辑 gaussian splatting NeRF scene reconstruction ✅
7 PanFlow: Decoupled Motion Control for Panoramic Video Generation PanFlow：解耦运动控制的全景视频生成方法 optical flow motion transfer ✅
8 Binary-Gaussian: Compact and Progressive Representation for 3D Gaussian Segmentation 提出Binary-Gaussian，用于压缩3D高斯分割的特征表示并提升分割精度。 3D gaussian splatting gaussian splatting
9 CircleFlow: Flow-Guided Camera Blur Estimation using a Circle Grid Target CircleFlow：利用圆形网格靶标和光流引导的相机模糊估计 optical flow localization
10 OmniFD: A Unified Model for Versatile Face Forgery Detection OmniFD：用于多功能人脸伪造检测的统一模型，提升效率和泛化性 localization ✅
11 Learning Eigenstructures of Unstructured Data Manifolds 提出一种直接从非结构化数据学习谱基的框架，用于形状和流形分析。 point cloud

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#题目一句话要点标签🔗⭐
12 Audio-Visual World Models: Towards Multisensory Imagination in Sight and Sound 提出AVWM框架，利用视听信息进行环境建模，提升智能体导航性能 world model localization navigation
13 S2AM3D: Scale-controllable Part Segmentation of 3D Point Cloud S2AM3D：提出可控粒度的三维点云部件分割方法 contrastive learning point cloud
14 Stronger is not better: Better Augmentations in Contrastive Learning for Medical Image Segmentation 针对医学图像分割，研究对比学习中更优的数据增强策略 representation learning contrastive learning
15 Accelerating Inference of Masked Image Generators via Reinforcement Learning 提出Speed-RL，通过强化学习加速掩码图像生成模型推理，显著减少采样步骤。 reinforcement learning

🔬 支柱一：机器人控制 (Robot Control) (4 篇)

#题目一句话要点标签🔗⭐
16 Silhouette-based Gait Foundation Model 提出FoundationGait，首个可扩展的步态自监督预训练框架，提升多种步态任务性能。 gait walking ✅
17 EmoDiffTalk:Emotion-aware Diffusion for Editable 3D Gaussian Talking Head EmoDiffTalk：提出情感感知扩散模型，用于可编辑的3D高斯说话头生成。 manipulation 3D gaussian splatting gaussian splatting
18 HanDyVQA: A Video QA Benchmark for Fine-Grained Hand-Object Interaction Dynamics HanDyVQA：一个用于细粒度手-物交互动态的视频问答基准 manipulation HOI
19 Charts Are Not Images: On the Challenges of Scientific Chart Editing 提出FigEdit基准，揭示现有生成模型在科学图表编辑中的结构化转换能力不足 manipulation ✅

🔬 支柱五：交互与反应 (Interaction & Reaction) (2 篇)

#题目一句话要点标签🔗⭐
20 Efficient and Scalable Monocular Human-Object Interaction Motion Reconstruction 提出4DHOISolver框架，结合人工标注，高效重建单目视频中的人-物交互运动。 human-object interaction HOI ✅
21 SocialFusion: Addressing Social Degradation in Pre-trained Vision-Language Models 提出SocialFusion框架，解决预训练视觉-语言模型中的社会认知退化问题 social interaction

⬅️ 返回 cs.CV 首页 · 🏠 返回主页