cs.CV（2025-02-20）

📊 共 21 篇论文 | 🔗 5 篇有代码

🎯 兴趣领域导航

支柱三：空间感知与语义 (Perception & Semantics) (7 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (7 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (4 🔗1) 支柱一：机器人控制 (Robot Control) (2) 支柱八：物理动画 (Physics-based Animation) (1)

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

#题目一句话要点标签🔗⭐
1 GS-Cache: A GS-Cache Inference Framework for Large-scale Gaussian Splatting Models GS-Cache：用于大规模高斯溅射模型的缓存加速推理框架 3D gaussian splatting 3DGS gaussian splatting
2 OrchardDepth: Precise Metric Depth Estimation of Orchard Scene from Monocular Camera Images OrchardDepth：单目相机果园场景精确度量深度估计 depth estimation monocular depth metric depth
3 Monocular Depth Estimation and Segmentation for Transparent Object with Iterative Semantic and Geometric Fusion 提出一种单目透明物体深度估计与分割的迭代语义几何融合框架 depth estimation monocular depth ✅
4 OG-Gaussian: Occupancy Based Street Gaussians for Autonomous Driving 提出OG-Gaussian，利用Occupancy Grid重建自动驾驶场景，降低成本并提升效率。 3D gaussian splatting 3DGS gaussian splatting
5 Learning Temporal 3D Semantic Scene Completion via Optical Flow Guidance FlowScene：利用光流引导的时序3D语义场景补全方法 optical flow
6 CrossOver: 3D Scene Cross-Modal Alignment CrossOver：提出一种灵活的跨模态3D场景对齐框架，用于解决多模态数据不完整和未对齐问题。 scene understanding
7 LXLv2: Enhanced LiDAR Excluded Lean 3D Object Detection with Fusion of 4D Radar and Camera 提出LXLv2以解决LiDAR排除的3D目标检测问题 occupancy grid

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

#题目一句话要点标签🔗⭐
8 Scaling Text-Rich Image Understanding via Code-Guided Synthetic Multimodal Data Generation CoSyn：利用代码引导的合成多模态数据生成，提升文本丰富图像理解能力 large language model multimodal
9 Multimodal RewardBench: Holistic Evaluation of Reward Models for Vision Language Models 提出Multimodal RewardBench，用于全面评估视觉语言模型奖励模型 multimodal ✅
10 Time Travel: A Comprehensive Benchmark to Evaluate LMMs on Historical and Cultural Artifacts 提出TimeTravel基准，用于评估LMMs在历史文化文物理解上的能力。 multimodal ✅
11 PLPHP: Per-Layer Per-Head Vision Token Pruning for Efficient Large Vision-Language Models 提出PLPHP以解决大规模视觉语言模型的推理效率问题 multimodal
12 Integrating Extra Modality Helps Segmentor Find Camouflaged Objects Well 提出MultiCOS框架，融合多模态信息提升伪装目标分割性能 multimodal ✅
13 Evaluating Precise Geolocation Inference Capabilities of Vision Language Models 评估视觉语言模型在精确地理位置推断方面的能力 foundation model
14 LLM-EvRep: Learning an LLM-Compatible Event Representation Using a Self-Supervised Framework 提出LLM-EvRep，利用自监督框架学习LLM兼容的事件表示，提升事件识别性能 large language model

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

#题目一句话要点标签🔗⭐
15 Self-supervised Monocular Depth Estimation Robust to Reflective Surface Leveraged by Triplet Mining 提出基于Triplet Mining的自监督单目深度估计方法，提升反射表面深度估计鲁棒性 distillation depth estimation monocular depth
16 Vision Foundation Models in Medical Image Analysis: Advances and Challenges 综述医学影像分析中视觉基础模型的研究进展与挑战，聚焦分割任务。 distillation foundation model
17 Structurally Disentangled Feature Fields Distillation for 3D Understanding and Editing 提出结构解耦特征场蒸馏方法，用于三维理解与编辑 distillation
18 LongWriter-V: Enabling Ultra-Long and High-Fidelity Generation in Vision-Language Models LongWriter-V：通过长文本SFT数据集和迭代DPO，实现视觉语言模型中的超长且高保真生成。 DPO direct preference optimization ✅

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
19 Designing Parameter and Compute Efficient Diffusion Transformers using Distillation 利用知识蒸馏设计参数和计算高效的Diffusion Transformer，适用于边缘设备。 Apple Vision Pro distillation
20 Fostering Inclusion: A Virtual Reality Experience to Raise Awareness of Dyslexia-Related Barriers in University Settings 提出一种基于VR的共情体验，旨在提升大学环境中对阅读障碍相关障碍的认知。 locomotion

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

#题目一句话要点标签🔗⭐
21 Daily Land Surface Temperature Reconstruction in Landsat Cross-Track Areas Using Deep Ensemble Learning With Uncertainty Quantification 提出DELAG深度集成学习方法，重建Landsat跨轨区域高时空分辨率地表温度，并量化不确定性。 spatiotemporal

⬅️ 返回 cs.CV 首页 · 🏠 返回主页