cs.CV（2026-06-29）

📊 共 58 篇论文 | 🔗 11 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (19 🔗2) 支柱三：空间感知与语义 (Perception & Semantics) (15 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (11 🔗2) 支柱一：机器人控制 (Robot Control) (5 🔗1) 支柱四：生成式动作 (Generative Motion) (3) 支柱七：动作重定向 (Motion Retargeting) (3 🔗3) 支柱六：视频提取与匹配 (Video Extraction) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (19 篇)

#题目一句话要点标签🔗⭐
1 Latent Noise Mask for Reducing Visual Redundancy in Multimodal Large Language Models 提出LatEnt Noise Mask以解决多模态大语言模型的视觉冗余问题 large language model multimodal chain-of-thought
2 LLM-based Multimodal Personality Recognition via Facial Action Unit-Text Semantic Fusion 提出基于LLM的多模态个性识别方法以解决信息损失问题 large language model multimodal
3 Clearer Sight, Fewer Lies: Oriented Pickup Preference Optimization for Multimodal Hallucination Mitigation 提出OPPO以优化多模态幻觉缓解问题 large language model multimodal
4 Unlocking the Visual Record of Materials Science: A Large-Scale Multimodal Dataset from Scientific Literature 提出MatMMExtract以解决材料科学文献中的视觉信息提取问题 large language model multimodal
5 Residual-Guided Expert Specialization for Incomplete Multimodal Learning 提出MARS框架以解决不完整多模态学习问题 multimodal
6 Illuminating Unified Multimodal Model for Free-form Interleaved Text-Image Generation 提出ILLUME-X以解决多模态文本-图像生成问题 multimodal
7 Uncertainty Estimation in Pathology Foundation Models via Deep Mutual Learning 提出DICE框架以解决病理基础模型的不确定性估计问题 foundation model
8 Traffic-CBM: A Structurally Interpretable Multimodal Framework for Encrypted Traffic Classification 提出Traffic-CBM以解决加密流量分类可解释性问题 multimodal
9 Benchmarking Geospatial Foundation Models for Agriculture Applications 提出地理空间基础模型基准以解决农业应用中的迁移性问题 foundation model
10 Orca: The World is in Your Mind 提出Orca模型以实现对世界状态的统一理解与预测 foundation model multimodal
11 Semantic-Driven Scale and Spatial Selection for Efficient Cross-Modal Alignment in Referring Remote Sensing Image Segmentation 提出S4ECA以解决遥感图像分割中的跨模态对齐问题 foundation model multimodal
12 MuseBench: Benchmarking Intent-Level Audiovisual Arts Understanding in MLLMs 提出MuseBench以解决多模态大语言模型在艺术理解中的不足问题 large language model multimodal
13 Rigel: Self-Distilled Score Adaptation for Image and Video Captioning Evaluation 提出Rigel以解决图像和视频字幕评估中的人类判断对齐问题 large language model multimodal
14 Latent-CURE for Breast Cancer Diagnosis 提出Latent-CURE以解决乳腺癌诊断中的偏差问题 multimodal chain-of-thought
15 See Only When Needed: Context-Aware Attention Intervention for Mitigating Hallucinations in LVLMs 提出上下文感知注意力干预以缓解LVLM中的幻觉问题 multimodal visual grounding ✅
16 UnfoldArt: Zero-Shot Recovery of Full Articulated 3D Objects from Text or Image 提出UnfoldArt以解决稀疏观测下的全关节3D物体重建问题 embodied AI
17 Learning from Reliable Latent Prompts for Visual Recognition with Missing Modalities 提出可靠潜在提示学习以解决缺失模态问题 multimodal
18 BrainJanus: A Unified Model for Understanding and Generation across Brain, Vision, and Language 提出BrainJanus以解决脑、视觉和语言之间的多模态建模问题 multimodal ✅
19 One Forward Beats Two: InnerZoom for Accurate and Efficient GUI Grounding 提出InnerZoom以解决GUI定位效率与准确性问题 instruction following

🔬 支柱三：空间感知与语义 (Perception & Semantics) (15 篇)

#题目一句话要点标签🔗⭐
20 StereoGS: Sparse-View 3D Gaussian Splatting via Stereo Priors 提出StereoGS以解决稀疏视图下3D高斯点云过拟合问题 depth estimation monocular depth stereo depth ✅
21 Open-Vocabulary and Referring Segmentation for 3D Gaussians Using 2D Detectors 提出GaussDet以解决3D高斯分割中的语言驱动问题 3D gaussian splatting 3DGS gaussian splatting
22 IBRSteG: Learning a Generalizable Steganography Framework for 3D Gaussian Splatting 提出IBRSteG以解决3D高斯点云隐写的通用性问题 3D gaussian splatting 3DGS gaussian splatting ✅
23 UniTriSplat: A Unified 3D Gaussian Splatting Framework with Uniform Spherical Rasterization for Universal Cameras 提出UniTriSplat以解决3D高斯点云在通用相机中的渲染问题 3D gaussian splatting 3DGS gaussian splatting
24 Graph-GSReg: Leveraging 3D Scene Graphs for Gaussian Splatting Registration 提出Graph-GSReg以解决3D高斯场景合并问题 3D gaussian splatting 3DGS gaussian splatting
25 OP3DSG: Open-Vocabulary Part-Aware 3D Scene Graph Generation for Real-World Environments 提出OP3DSG以解决现有3D场景图生成的局限性 open-vocabulary open vocabulary affordance
26 Shell-Supervised Gaussian Splatting for Urban Real-to-Sim Reconstruction 提出壳监督高斯点云重建以解决城市近景重建问题 3D reconstruction gaussian splatting splatting
27 FastPano3D: Feed-Forward Indoor Panoramic 3D Reconstruction from a Single Image 提出FastPano3D以解决单图像快速室内三维重建问题 3DGS 3D reconstruction gaussian splatting
28 Learning Efficient 4D Gaussian Representations from Monocular Videos with Flow Splatting 提出Flow Splatting以解决动态场景重建效率问题 3D gaussian splatting gaussian splatting splatting
29 Robust and Efficient Monocular 3D Gaussian SLAM for Kilometer-Scale Outdoor Scenes 提出KiloGS-SLAM以解决长距离户外场景的跟踪与内存问题 3D gaussian splatting 3DGS gaussian splatting
30 AerialMetric: Benchmarking and Adapting UAV Monocular Metric Depth Estimation in the Real World 提出AerialMetric以解决无人机单目度量深度估计问题 depth estimation metric depth ✅
31 Argus: Metric Panoramic 3D Reconstruction for Indoor Scenes 提出Argus以解决室内场景的度量全景3D重建问题 depth estimation 3D reconstruction geometric consistency
32 Efficient Visual Pointing for Embodied AI:Agent-Driven Data Synthesis, Cross-Block Attention, and Iterative Correction 提出高效视觉指向方法以解决嵌入式AI的准确性问题 affordance embodied AI
33 SADL: What to Ignore? A Benchmark for Subject-Aware Distractor Localization 提出SADL基准以解决主观意识下的干扰物定位问题 open-vocabulary open vocabulary multimodal
34 MUSE: Unlocking Timestep as Native Task Steering for One-Step Dense Prediction 提出MUSE以解决单步稠密预测中的多任务学习问题 monocular depth

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

#题目一句话要点标签🔗⭐
35 Monte Carlo Energy Aggregation for Mobile 3D Gaussian Splatting 提出Flux-GS以解决移动平台高效3D高斯渲染问题 distillation 3D gaussian splatting gaussian splatting ✅
36 OmniDance: Multimodal Driven Dance Video Generation with Large-scale Internet Data 提出OmniDance以解决舞蹈视频生成中的数据不足与多模态融合问题 curriculum learning human motion foundation model ✅
37 OWMDrive: Causality-Aware End-to-End Autonomous Driving via 4D Occupancy World Model 提出OWMDrive以解决复杂交通环境下的自主驾驶决策问题 world model world models spatiotemporal
38 Consistency as Inductive Bias: Learning Cross-View Invariance for Robust Multimodal Reasoning 提出ConsistRoll以解决多模态推理中的跨视图一致性问题 reinforcement learning multimodal
39 LWDrive: Layer-Wise World-Model-Guided Vision-Language Model Planning for Autonomous Driving 提出LWDrive以解决自主驾驶规划中的轨迹精度不足问题 world model world models
40 Cross-Modal Iteration Distillation for Robust IHD Screening: The IDNet Framework and A New Benchmark 提出IDNet框架以解决缺乏基准和多模态融合不足的问题 distillation multimodal
41 SUMO: Segment and Track Any Motion with Nonlinear State Space Models 提出SUMO框架以解决复杂运动对象的跟踪与分割问题 SSM state space model
42 Variance Reduction on the Camera Axis: Multi-View Score Distillation for 3D 提出多视角得分蒸馏以降低3D生成中的方差问题 distillation
43 H-GRPO: Permutation-Invariant Reinforcement Learning for Grounded Visual Reasoning 提出H-GRPO框架以解决视觉推理中的可解释性问题 reinforcement learning
44 Robust Trajectory Distillation: Hybrid Reweighting Meets Teacher-Inspired Targets 提出基于轨迹蒸馏的框架以解决噪声监督下的数据集蒸馏问题 distillation
45 HomeDiffusion: Zero-Shot Object Customization with Multi-View Representation Learning for Indoor Scenes 提出HomeDiffusion以解决室内场景中物体定制生成问题 representation learning

🔬 支柱一：机器人控制 (Robot Control) (5 篇)

#题目一句话要点标签🔗⭐
46 Early Warning Signals for OpenVLA Failure under Visual Distribution Shift 提出监测机制以应对OpenVLA在视觉分布变化下的失败预警 manipulation vision-language-action OpenVLA
47 OmniCoT: A Benchmark for Global and Multi-Step Panoramic Reasoning 提出OmniCoT以解决全局多步全景推理问题 sim-to-real large language model multimodal
48 Towards in-the-wild Egocentric 3D Hand-Object Pose Estimation 提出EPIC-Contact数据集与HOPformer模型以解决野外环境下的手-物体3D姿态估计问题 bi-manual egocentric ✅
49 Early Cue Precision Shapes Visual Shortcut Learning in Controlled Cue-Manipulation Benchmarks 提出早期线索精度以改善视觉快捷学习问题 manipulation
50 Goku: A Million-Scale Universal Dataset and Benchmark for Instruction-Based Video Editing 提出Goku数据集以解决复杂视频编辑任务的不足问题 manipulation instruction following

🔬 支柱四：生成式动作 (Generative Motion) (3 篇)

#题目一句话要点标签🔗⭐
51 PoseShield: Neural Collision Fields for Human Self-Collision Resolution 提出PoseShield以解决人类自碰撞问题 motion synthesis motion generation penetration
52 SICAGE: Speaker-Independent Culture-Aware Gesture Generation using TED4C-L Dataset 提出SICAGE框架以解决文化差异对手势生成的影响 motion synthesis multimodal
53 DCGrasp: Distance-aware Controllable Grasp Generation 提出DCGrasp以解决3D手-物体交互生成问题 physically plausible

🔬 支柱七：动作重定向 (Motion Retargeting) (3 篇)

#题目一句话要点标签🔗⭐
54 3D Scene-Adaptive Trajectory-Controllable Human Image Animation with Camera Movement 提出场景自适应的人物图像动画框架以解决摄像机视角变化问题 motion retargeting human motion foundation model ✅
55 CylindTrack: Depth-Aware Cylindrical Motion Modeling for Panoramic Multi-Object Tracking 提出CylindTrack以解决全景多目标跟踪中的深度感知问题 motion prediction ✅
56 Scene-aware Prediction of Diverse Human Movement Goals 提出基于生成模型的多目标预测方法以解决人类运动目标的随机性问题 human motion ✅

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

#题目一句话要点标签🔗⭐
57 The Surprising Effectiveness of Video Diffusion Models for Hand Motion Reconstruction 提出ViDiHand以解决手部运动重建中的遮挡问题 egocentric motion reconstruction embodied AI

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#题目一句话要点标签🔗⭐
58 Reweighting Framewise Attention in Video Transformers for Facial Expression Understanding 提出MiRA框架以解决视频中面部表情理解的注意力不足问题 spatiotemporal

⬅️ 返回 cs.CV 首页 · 🏠 返回主页