cs.CV（2025-01-06）

📊 共 26 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) (14 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (4 🔗1) 支柱六：视频提取与匹配 (Video Extraction) (2) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (14 篇)

#题目一句话要点标签🔗⭐
1 EAGLE: Enhanced Visual Grounding Minimizes Hallucinations in Instructional Multimodal Models EAGLE：增强视觉基础能力，最小化指令型多模态模型中的幻觉问题 large language model multimodal visual grounding
2 Socratic Questioning: Learn to Self-guide Multimodal Reasoning in the Wild 提出Socratic Questioning框架，提升多模态LLM在复杂视觉推理中的性能。 large language model multimodal chain-of-thought
3 CM3T: Framework for Efficient Multimodal Learning for Inhomogeneous Interaction Datasets CM3T：一种高效的多模态学习框架，用于异构交互数据集。 multimodal
4 MObI: Multimodal Object Inpainting Using Diffusion Models MObI：提出基于扩散模型的多模态物体填充框架，用于自动驾驶场景数据增强。 multimodal
5 MVP: Multimodal Emotion Recognition based on Video and Physiological Signals 提出MVP模型，融合视频与生理信号，提升长时序情感识别性能 multimodal
6 FoundPAD: Foundation Models Reloaded for Face Presentation Attack Detection FoundPAD：利用重载的基础模型进行人脸呈现攻击检测 foundation model ✅
7 Large Language Models for Video Surveillance Applications 提出基于视觉语言模型的视频监控摘要生成方法，提升分析精度和效率。 large language model
8 Visual Large Language Models for Generalized and Specialized Applications 综述视觉大语言模型在通用和专用场景下的应用，并探讨其挑战与未来方向 large language model ✅
9 Ultrasound-QBench: Can LLMs Aid in Quality Assessment of Ultrasound Imaging? Ultrasound-QBench：利用多模态大语言模型辅助超声图像质量评估 large language model multimodal
10 SAM-EM: Real-Time Segmentation for Automated Liquid Phase Transmission Electron Microscopy SAM-EM：用于自动化液相透射电子显微镜的实时分割方法 foundation model ✅
11 CAT: Content-Adaptive Image Tokenization 提出内容自适应图像Token化方法CAT，提升图像重建和生成效果。 large language model
12 SceneVTG++: Controllable Multilingual Visual Text Generation in the Wild SceneVTG++：提出可控多语言场景视觉文本生成方法，解决自然场景图像文本生成难题 large language model
13 MDP3: A Training-free Approach for List-wise Frame Selection in Video-LLMs 提出MDP3以解决视频大语言模型中的帧选择问题 large language model
14 Found in Translation: semantic approaches for enhancing AI interpretability in face verification 提出基于语义概念的XAI框架，提升人脸验证模型的可解释性 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (5 篇)

#题目一句话要点标签🔗⭐
15 Gaussian Masked Autoencoders 提出高斯掩码自编码器(GMAE)，联合学习语义抽象和空间理解，实现零样本空间理解能力。 representation learning masked autoencoder MAE ✅
16 AE-NeRF: Augmenting Event-Based Neural Radiance Fields for Non-ideal Conditions and Larger Scene AE-NeRF：增强事件相机NeRF在非理想条件和更大场景下的重建能力 distillation NeRF neural radiance field
17 Human Gaze Boosts Object-Centered Representation Learning 提出基于人类注视机制的物体中心表征学习方法，提升自监督学习性能 representation learning egocentric Ego4D
18 First-place Solution for Streetscape Shop Sign Recognition Competition 提出多阶段融合框架，解决复杂街景店招文字识别难题 reinforcement learning multimodal
19 CCStereo: Audio-Visual Contextual and Contrastive Learning for Binaural Audio Generation 提出基于音视频上下文对比学习的双耳音频生成模型，提升空间细节表现。 contrastive learning

🔬 支柱三：空间感知与语义 (Perception & Semantics) (4 篇)

#题目一句话要点标签🔗⭐
20 Compression of 3D Gaussian Splatting with Optimized Feature Planes and Standard Video Codecs 提出基于优化特征平面和标准视频编解码器的3D高斯溅射压缩方法 3D gaussian splatting gaussian splatting splatting ✅
21 Spiking monocular event based 6D pose estimation for space application 提出基于脉冲神经网络的单目事件相机6D位姿估计方法，用于空间应用 6D pose estimation
22 Pointmap-Conditioned Diffusion for Consistent Novel View Synthesis PointmapDiff：利用点云图条件扩散模型实现一致性新视角合成 3D gaussian splatting gaussian splatting splatting
23 ProTracker: Probabilistic Integration for Robust and Accurate Point Tracking ProTracker：融合概率的鲁棒精确视频点跟踪框架 optical flow

🔬 支柱六：视频提取与匹配 (Video Extraction) (2 篇)

#题目一句话要点标签🔗⭐
24 HaWoR: World-Space Hand Motion Reconstruction from Egocentric Videos HaWoR：提出一种从第一视角视频重建世界坐标系下手部运动的高保真方法 egocentric hand reconstruction
25 WorldPose: A World Cup Dataset for Global 3D Human Pose Estimation WorldPose：提出世界杯多视角3D人体姿态估计数据集，挑战现有算法。 SMPL

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
26 HOGSA: Bimanual Hand-Object Interaction Understanding with 3D Gaussian Splatting Based Data Augmentation 提出基于3D高斯溅射的数据增强框架，用于提升双手动目标交互理解 bi-manual 3D gaussian splatting 3DGS

⬅️ 返回 cs.CV 首页 · 🏠 返回主页