cs.CV（2025-08-12）

📊 共 29 篇论文 | 🔗 5 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (9 🔗2) 支柱三：空间感知与语义 (Perception & Semantics) (8) 支柱二：RL算法与架构 (RL & Architecture) (4 🔗1) 支柱八：物理动画 (Physics-based Animation) (3 🔗2) 支柱四：生成式动作 (Generative Motion) (3) 支柱五：交互与反应 (Interaction & Reaction) (1) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (9 篇)

#题目一句话要点标签🔗⭐
1 ColorGPT: Leveraging Large Language Models for Multimodal Color Recommendation 提出ColorGPT以解决多模态颜色推荐问题 large language model multimodal
2 Spatial Traces: Enhancing VLA Models with Spatial-Temporal Understanding 提出空间轨迹方法以增强VLA模型的时空理解能力 vision-language-action VLA ✅
3 3DFroMLLM: 3D Prototype Generation only from Pretrained Multimodal LLMs 提出3DFroMLLM以解决多模态大语言模型空间推理不足问题 large language model multimodal
4 DiffPose-Animal: A Language-Conditioned Diffusion Framework for Animal Pose Estimation 提出DiffPose-Animal以解决动物姿态估计问题 large language model language conditioned
5 Harnessing Input-Adaptive Inference for Efficient VLN 提出输入自适应推理方法以提升视觉语言导航效率 VLN ✅
6 MMIF-AMIN: Adaptive Loss-Driven Multi-Scale Invertible Dense Network for Multimodal Medical Image Fusion 提出MMIF-AMIN以解决多模态医学图像融合问题 multimodal
7 MADPromptS: Unlocking Zero-Shot Morphing Attack Detection with Multiple Prompt Aggregation 提出MADPromptS以解决零-shot人脸变形攻击检测问题 foundation model multimodal
8 Scaling Learned Image Compression Models up to 1 Billion 提出大规模学习图像压缩模型以提升压缩性能 large language model
9 TaoCache: Structure-Maintained Video Generation Acceleration 提出TaoCache以解决视频生成加速中的结构一致性问题 instruction following

🔬 支柱三：空间感知与语义 (Perception & Semantics) (8 篇)

#题目一句话要点标签🔗⭐
10 Gradient-Direction-Aware Density Control for 3D Gaussian Splatting 提出梯度方向感知密度控制以解决3D高斯点云渲染问题 3D gaussian splatting 3DGS gaussian splatting
11 GaussianUpdate: Continual 3D Gaussian Splatting Update for Changing Environments 提出GaussianUpdate以解决动态环境下的3D场景更新问题 3D gaussian splatting gaussian splatting splatting
12 Vision-Only Gaussian Splatting for Collaborative Semantic Occupancy Prediction 提出稀疏3D语义高斯点云以解决协作语义占用预测问题 depth estimation gaussian splatting splatting
13 A new dataset and comparison for multi-camera frame synthesis 提出多摄像头数据集以解决帧合成方法比较问题 depth estimation 3D gaussian splatting gaussian splatting
14 MonoPartNeRF:Human Reconstruction from Monocular Video via Part-Based Neural Radiance Fields 提出MonoPartNeRF以解决单目视频中人类重建问题 NeRF neural radiance field
15 HQ-OV3D: A High Box Quality Open-World 3D Detection Framework based on Diffision Model 提出HQ-OV3D以解决开放世界3D检测中的伪标签质量问题 open-vocabulary open vocabulary geometric consistency
16 ViPE: Video Pose Engine for 3D Geometric Perception 提出ViPE以解决3D几何感知中的视频标注挑战 metric depth
17 Revisiting Efficient Semantic Segmentation: Learning Offsets for Better Spatial and Class Feature Alignment 提出偏移学习方法以解决语义分割中的特征对齐问题 scene understanding

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

#题目一句话要点标签🔗⭐
18 DocThinker: Explainable Multimodal Large Language Models with Rule-based Reinforcement Learning for Document Understanding 提出DocThinker以解决多模态大语言模型的可解释性与适应性问题 reinforcement learning policy learning large language model ✅
19 UltraLight Med-Vision Mamba for Classification of Neoplastic Progression in Tubular Adenomas 提出Ultralight Med-Vision Mamba以解决肠道腺瘤分类问题 Mamba SSM
20 Addressing Bias in VLMs for Glaucoma Detection Without Protected Attribute Supervision 提出无监督属性去偏见方法以改善青光眼检测 contrastive learning multimodal
21 AME: Aligned Manifold Entropy for Robust Vision-Language Distillation 提出AME以解决视觉-语言蒸馏中的不确定性问题 distillation

🔬 支柱八：物理动画 (Physics-based Animation) (3 篇)

#题目一句话要点标签🔗⭐
22 FusionEnsemble-Net: An Attention-Based Ensemble of Spatiotemporal Networks for Multimodal Sign Language Recognition 提出FusionEnsemble-Net以解决多模态手语识别问题 spatiotemporal multimodal ✅
23 KFFocus: Highlighting Keyframes for Enhanced Video Understanding 提出KFFocus以解决视频理解中的关键帧压缩问题 spatiotemporal large language model multimodal
24 UniConvNet: Expanding Effective Receptive Field while Maintaining Asymptotically Gaussian Distribution for ConvNets of Any Scale 提出UniConvNet以扩展有效感受野并保持高斯分布 UniCon ✅

🔬 支柱四：生成式动作 (Generative Motion) (3 篇)

#题目一句话要点标签🔗⭐
25 X-UniMotion: Animating Human Images with Expressive, Unified and Identity-Agnostic Motion Latents 提出X-UniMotion以实现高保真、身份无关的人体动画 motion latent
26 Spatial-Temporal Multi-Scale Quantization for Flexible Motion Generation 提出多尺度量化方法以解决人类动作生成的灵活性问题 motion generation
27 RealisMotion: Decomposed Human Motion Control and Video Generation in the World Space 提出RealisMotion以解决人类运动控制与视频生成的挑战 text-to-motion

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#题目一句话要点标签🔗⭐
28 QueryCraft: Transformer-Guided Query Initialization for Enhanced Human-Object Interaction Detection 提出QueryCraft以解决HOI检测中查询初始化不足问题 human-object interaction HOI

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
29 SegDAC: Improving Visual Reinforcement Learning by Extracting Dynamic Objectc-Centric Representations from Pretrained Vision Models 提出SegDAC以解决视觉强化学习中的动态对象表示问题 manipulation reinforcement learning

⬅️ 返回 cs.CV 首页 · 🏠 返回主页