cs.CV（2026-10-21）

📊 共 38 篇论文 | 🔗 8 篇有代码

🎯 兴趣领域导航

支柱三：空间感知与语义 (Perception & Semantics) (12 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (11 🔗2) 支柱一：机器人控制 (Robot Control) (7 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (7 🔗1) 支柱六：视频提取与匹配 (Video Extraction) (1)

🔬 支柱三：空间感知与语义 (Perception & Semantics) (12 篇)

#题目一句话要点标签🔗⭐
1 AOV-GS: Active Open-Vocabulary Semantic Mapping with 3D Gaussian Splatting 提出AOV-GS以解决开放词汇语义映射问题 3D gaussian splatting gaussian splatting splatting ✅
2 ClunoGS: Compact Language 3D Gaussian Splatting via Unified Optimization 提出ClunoGS以解决语言3D高斯点云压缩问题 3D gaussian splatting 3DGS gaussian splatting
3 OpenEarth-OVS44: A Multimodal Remote Sensing Dataset for Open-Vocabulary Segmentation 提出OpenEarth-OVS44以解决遥感开放词汇分割问题 open-vocabulary open vocabulary multimodal
4 Open-Vocabulary Models and Label Propagation for Uncertainty-aware Pseudo-Labeling of Object Detection Datasets 提出半监督框架以解决目标检测数据集的不确定性标注问题 open-vocabulary open vocabulary
5 LLMSeg: Training-Free Text-Prompted Image Segmentation with B-Splines 提出LLMSeg以实现无训练的文本提示图像分割 open-vocabulary open vocabulary large language model
6 AutoPhyX: Automatic Text-Condition Physics Property Generation 提出AutoPhyX以解决物理属性生成的模糊性问题 gaussian splatting splatting NeRF
7 SICAF: Time-Varying Focus Bottleneck for Self-Supervised Event-Based Optical Flow with Spiking Neural Network 提出SICAF以解决自监督事件驱动光流估计中的时序特征提取问题 optical flow spatiotemporal
8 DriftNeRF: Cross-Ray Density Drifting for Sparse-View Neural Radiance Fields 提出DriftNeRF以解决稀疏视图下NeRF的密度漂移问题 NeRF neural radiance field
9 A Unified Evaluation Framework of Physical Adversarial Robustness in the Edge VLM Era 提出统一评估框架以研究边缘VLM的物理对抗鲁棒性 open-vocabulary open vocabulary visual grounding ✅
10 Extending Affordance Segmentation for Wearable Robots to Multi-Object Images 提出一种改进方法以解决多物体场景下的抓取识别问题 affordance
11 BlazeVGGT: 13x Faster VGGT at 2,000 Frames via Camera Token without Training 提出BlazeVGGT以解决VGGT在长序列推理中的效率问题 VGGT ✅
12 Failure Modes of Vision-Language Models for Human Activity Understanding in Thermal HRI 评估热成像人机交互中的视觉语言模型失败模式 scene understanding

🔬 支柱二：RL算法与架构 (RL & Architecture) (11 篇)

#题目一句话要点标签🔗⭐
13 DewHand: Direct Metric World-Space Hand Motion Reconstruction from Egocentric Videos 提出DewHand以解决从自我中心视频重建手部运动的问题 flow matching VGGT motion generation
14 ALPS: Visual State Encoders for World Models 提出ALPS以统一视觉状态编码解决世界模型的多重任务问题 world model world models JEPA
15 Generalized CycleGRPO for Fine-Grained Multimodal Understanding and Grounding 提出Generalized CycleGRPO以解决细粒度多模态理解与定位问题 reinforcement learning large language model multimodal ✅
16 MS2CAST: A Video Anomaly Detection Framework via Multi-Scale Spatial Cross-Attention and Efficient Spatiotemporal Networks 提出MS2CAST框架以解决视频异常检测中的计算复杂性问题 Mamba optical flow spatiotemporal
17 APED-CXR: Adaptive Privileged Evidence Distillation for Chest X-Ray Grounding with Image-Only Inference 提出APED-CXR以解决胸部X光图像定位问题 distillation multimodal
18 Unlocking Spatial Grounding in Flow-Matching Models via SNR Tuning 通过SNR调优解锁流匹配模型中的空间定位能力 flow matching foundation model
19 Supervision-Adaptive Curriculum and Guarded Anti-Curriculum Learning for Dense Prediction 提出PaceShift以解决密集预测中的监督可靠性问题 curriculum learning foundation model
20 AdvFidelity: Unifying Preference Alignment and Few-Step Image Generation within Reinforcement Learning 提出AdvFidelity以解决图像生成中的偏好对齐与高效性问题 reinforcement learning distillation
21 OccJEPA: Occlusion as Masking for Amodal Video Segmentation 提出OccJEPA以解决视频遮挡下的物体分割问题 world model world models JEPA ✅
22 Enhancing Cross-Grade Generalization in Banana Variety Classification via Hybrid CNN-Transformers and Knowledge Self-Distillation 提出混合CNN-Transformer框架以解决香蕉品种跨等级分类问题 distillation
23 Looking Ahead: Modeling Future Evolution in Autoregressive Video Generation 提出未来块预测以解决自回归视频生成中的场景演变问题 world model world models

🔬 支柱一：机器人控制 (Robot Control) (7 篇)

#题目一句话要点标签🔗⭐
24 GeoMem-VLA: Internalizing 3D Scene Understanding into Temporal Memory for Vision-Language-Action Models 提出GeoMem-VLA以解决机器人操作中的3D场景理解问题 manipulation flow matching 3D reconstruction
25 CausalAlign: Causal Alignment of Video Generation Models for Consistent Ego-Motion 提出CausalAlign以解决视频生成模型中的摄像机自运动一致性问题 manipulation mobile manipulation world action model
26 UniRAP: Towards Unified Part-level Physical Affordance Reasoning and Actionable Perception 提出UniRAP以解决部分物理可供性推理与可操作感知问题 manipulation affordance spatiotemporal
27 Rethinking Video Autoencoders for Interactive World Models 提出CADRE以优化视频自编码器在交互式世界模型中的应用 manipulation world model world models
28 LiveAct: Towards Real-Time Multi-Agent World Models with Heterogeneous Action Interfaces 提出LiveAct以解决多代理异构动作接口的实时建模问题 manipulation world model world models
29 NeuralSRender: G-Buffer Guided Video Diffusion for Sim-to-Real Rendering 提出NeuralSRender以解决合成视频真实感不足问题 sim-to-real geometric consistency ✅
30 UNIVERS: Learning UNIversal Voxel-wise Embedding via Randomized Synthesis for generalizable one-shot landmark localization 提出UNIVERS以解决多模态体积扫描中的一-shot标志物定位问题 domain randomization contrastive learning ✅

🔬 支柱九：具身大模型 (Embodied Foundation Models) (7 篇)

#题目一句话要点标签🔗⭐
31 Balancing Foundation Models and Trait-Specific Learning for Forehead-Crease Recognition 提出基于额头皱纹识别的生物特征识别方法以应对部分面部可见性问题 large language model foundation model multimodal
32 Asking a Multimodal LLM Whether Someone Fell: Zero-Shot Screening of Low-Resolution Thermal Video in a Laboratory and an Occupied Home 提出多模态LLM以实现低分辨率热视频的跌倒检测 large language model multimodal
33 Plausible but Unimplantable: A Trustworthiness Audit of Shape Completion Foundation Models for Cranial Defect Reconstruction 审计颅骨缺损重建的形状补全基础模型的可信度 foundation model
34 Melange: A Synthetic Visual Bias Benchmark for Multi-modal Large Language Models 提出Melange基准以评估多模态大语言模型的社会偏见 large language model
35 HumanCentric-Scenes-AI: A Human-Curated Multimodal Dataset of AI-Generated Human-Centric Scenes Across Surveillance, Occupational Portraits, AI-Edited Images, and Short-Form Video 提出HumanCentric-Scenes-AI数据集以促进人类感知研究 multimodal
36 LocalAgent: Collaborative Agentic Verification for Fine-Grained Instance-Level Consistency 提出LocalAgent以解决细粒度实例级一致性验证问题 large language model multimodal ✅
37 From Canon to Nikon: Camera Domain Shift and Potential Equity Implications in Retinal AI 评估相机域转移对视网膜AI性能的影响 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🔗⭐
38 ProactEgoMem: Benchmarking Proactive Memory in Streaming Egocentric Video 提出ProactEgoMem以评估流媒体自我中心视频中的主动记忆 egocentric multimodal

⬅️ 返回 cs.CV 首页 · 🏠 返回主页