cs.CV（2026-08-12）

📊 共 34 篇论文 | 🔗 7 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (13 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (8 🔗2) 支柱三：空间感知与语义 (Perception & Semantics) (7 🔗2) 支柱七：动作重定向 (Motion Retargeting) (3) 支柱一：机器人控制 (Robot Control) (3 🔗1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (13 篇)

#题目一句话要点标签🔗⭐
1 Context Blindness in DPO: Mitigating Object Hallucination in MLLMs via Context-Calibrated Preference Optimization 提出C²-DPO以解决多模态大语言模型中的物体幻觉问题 DPO direct preference optimization large language model ✅
2 SCOUT: Unlocking Enhanced Spatial Reasoning via Structured Chain-of-Thought and Multi-Objective Process Reward 提出SCOUT以解决视觉语言模型空间推理不足问题 reinforcement learning chain-of-thought
3 GeoBridge: Decoupled Semantic Conditioning for Generative Image Geolocalization 提出GeoBridge以解决图像地理定位中的语义条件问题 flow matching large language model multimodal
4 EGM-Det: Entropy-Guided Multimodal Adaptive Fusion for UAV RGB-IR Object Detection 提出EGM-Det以解决UAV RGB-IR目标检测中的模态融合问题 distillation multimodal
5 Better Slots, Better Worlds: Representation Quality & Robustness in Object-Centric World Models 提出改进的对象中心世界模型以提升规划能力与鲁棒性 world model world models
6 How Can Driving World Models Do Counterfactual Prediction? 提出新方法以解决驾驶世界模型中的反事实预测问题 world model world models
7 Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction 提出GAS框架以解决多模态理解与生成目标不一致问题 representation learning large language model multimodal
8 STAR: A Spatial-Topology Aware Routing Framework for Generalizable 3D Scene Understanding 提出STAR框架以解决多模态3D场景理解中的拓扑差异问题 representation learning scene understanding ✅
9 HarmoniDPO: Video-guided Audio Generation via Preference-Optimized Diffusion 提出HarmoniDPO以解决视频引导音频生成中的同步与质量问题 reinforcement learning RLHF DPO
10 Avatar-Forever: Decoupled Parallel Training for High-Quality Real-Time Infinite Avatars 提出Avatar-Forever以解决长视频生成中的质量与效率问题 distillation foundation model
11 Anti-Shortcut Distillation via Temporal Negative Knowledge Transfer 提出反快捷蒸馏方法以提升知识蒸馏效果 distillation
12 GeoFlow: Efficient Driving Video Generation via Geometry-Aligned Priors 提出GeoFlow以解决驾驶视频生成效率低下问题 flow matching spatiotemporal
13 TGRHuman: Text-Guided Realistic 3D Human Generation via Diffusion Renderer 提出TGRHuman以解决高质量3D人类生成问题 distillation NeRF

🔬 支柱九：具身大模型 (Embodied Foundation Models) (8 篇)

#题目一句话要点标签🔗⭐
14 Class Activation Mapping in Explainable Computer Vision: A Method-Centered Review of CNN, Transformer, and Foundation-Model-Era Visual Explanations 综述类激活映射方法以提升可解释计算机视觉 foundation model
15 Understanding Why Foundation Models Work for Diffusion-Generated Image Detection 探讨基础模型在扩散生成图像检测中的有效性 foundation model
16 Do You See What You Draw? A Semantic Closed-Loop Framework for Holistic Evaluation of Unified Multimodal Models 提出自生成理解框架以解决统一多模态模型评估问题 multimodal
17 Advancing MLLM-based UAV Image Understanding and Reasoning: A Benchmark and a Training-Free Multi-Agent System 提出UAV-MAS以解决无人机图像理解与推理问题 large language model multimodal visual grounding
18 DreamFly: Causal Memory and Receding-Horizon Diffusion Planning for Aerial Vision-Language Navigation 提出DreamFly以解决空中视觉语言导航中的历史上下文不足问题 VLA VLN
19 Beyond Trial-and-Error: Agentic Optimization for Image-to-Video Adherence 提出Agentic Self-Improvement框架以解决I2V模型控制不足问题 large language model multimodal
20 Diagram-MMU: A Multi-Modal Benchmark for Scientific Diagrams 提出Diagram-MMU基准以评估科学图表解析能力 large language model multimodal ✅
21 JieZi: A Large-Scale Expert-Audited Dataset and Benchmark for Ancient Chinese Character Exegesis 提出JieZi数据集以解决古代汉字释义研究的不足问题 large language model multimodal ✅

🔬 支柱三：空间感知与语义 (Perception & Semantics) (7 篇)

#题目一句话要点标签🔗⭐
22 Learning from Multimodal Pseudo-Labels for Robust Open-Vocabulary Instance and Panoptic Segmentation 提出多模态伪标签学习框架以解决开放词汇实例和全景分割问题 open-vocabulary open vocabulary multimodal
23 Zero-OVCD: Bridging Training-Free Foundation Models and Pseudo-Label Learning for Open-Vocabulary Change Detection 提出Zero-OVCD以解决开放词汇变化检测中的标注依赖问题 open-vocabulary open vocabulary foundation model ✅
24 Repurposing RGB-based Foundation Model for Depth Estimation on Thermal Images Using Hierarchical Supervision 提出RGB-HS框架以解决热图像深度估计问题 depth estimation foundation model
25 Seed2GS: Camera-Free, Training-Free Object Extraction from 3D Gaussian Scenes via a Single Reference-View Grounding 提出Seed2GS以解决无摄像机、无训练的3D物体提取问题 3D gaussian splatting 3DGS gaussian splatting
26 Map-Det3D: Metric Feed-Forward 3D Reconstruction Prior for Multi-view 3D Object Detection from Streaming Inputs 提出Map-Det3D以解决单目视频中的3D物体检测问题 3D reconstruction ✅
27 ScaleVid: Geometry-Aware Video Object Scaling with Mesh-Free Inference 提出ScaleVid以解决视频对象缩放中的几何一致性问题 3D reconstruction geometric consistency
28 Surfsvr: 2D Surface Priors as 3D Geometric Regularizers for Sparse Voxel Reconstruction 提出SurfSVR以解决稀疏体素重建中的几何碎片化问题 monocular depth

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

#题目一句话要点标签🔗⭐
29 Autonomous Telerehabilitation via Skeletal Motion Prediction and Joint-Level Performance Assessment 提出基于骨骼运动预测的自主远程康复系统以解决缺乏持续监督的问题 human motion motion prediction
30 HSTGFormer: Hyper Spatial-Temporal Graph Transformer for 3D Human Pose Estimation 提出HSTGFormer以解决3D人类姿态估计中的时空信息分离问题 human motion
31 PolarSym: Polar Geometry-aware Attention for CAD Floorplan Parsing 提出PolarSym以解决CAD平面解析中的几何对称性问题 spatial relationship

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
32 Motion-as-Prompt: Enhancing Motion Reasoning in Multimodal Large Language Models via Motion-Guided Cross-Frame Visual Prompting 提出Motion-as-Prompt以解决多模态大语言模型的运动推理问题 manipulation large language model multimodal ✅
33 AVA-Encoder: Towards Agent-Native Video Representation Learning 提出AVA-Encoder以解决代理智能视频表示学习问题 manipulation representation learning
34 A Hybrid Framework of Vision Transformer and Gated Recurrent Unit for Detection of Mosquito Diseases 提出混合框架结合视觉变换器与门控循环单元以检测蚊子疾病 locomotion

⬅️ 返回 cs.CV 首页 · 🏠 返回主页