cs.CV（2024-03-28）

📊 共 33 篇论文 | 🔗 13 篇有代码

🎯 兴趣领域导航

支柱三：空间感知与语义 (Perception & Semantics) (13 🔗5) 支柱二：RL算法与架构 (RL & Architecture) (8 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (5 🔗2) 支柱四：生成式动作 (Generative Motion) (3 🔗2) 支柱一：机器人控制 (Robot Control) (2 🔗1) 支柱七：动作重定向 (Motion Retargeting) (1) 支柱六：视频提取与匹配 (Video Extraction) (1 🔗1)

🔬 支柱三：空间感知与语义 (Perception & Semantics) (13 篇)

#题目一句话要点标签🔗⭐
1 GauStudio: A Modular Framework for 3D Gaussian Splatting and Beyond 提出GauStudio框架以提升3D高斯点云建模与渲染能力 3D gaussian splatting 3DGS gaussian splatting
2 FlowDepth: Decoupling Optical Flow for Self-Supervised Monocular Depth Estimation 提出FlowDepth以解决动态物体引起的深度估计不匹配问题 depth estimation monocular depth optical flow
3 CoherentGS: Sparse Novel View Synthesis with Coherent 3D Gaussians 提出CoherentGS以解决稀疏视图合成问题 monocular depth 3D gaussian splatting 3DGS
4 IVLMap: Instance-Aware Visual Language Grounding for Consumer Robot Navigation 提出IVLMap以解决实例级和属性级导航问题 semantic mapping semantic map VLMAP ✅
5 SA-GS: Scale-Adaptive Gaussian Splatting for Training-Free Anti-Aliasing 提出SA-GS以解决高频场景下的抗锯齿问题 gaussian splatting splatting ✅
6 TOGS: Gaussian Splatting with Temporal Opacity Offset for Real-Time 4D DSA Rendering 提出TOGS以解决4D DSA渲染质量与速度不足的问题 gaussian splatting splatting ✅
7 OV-Uni3DETR: Towards Unified Open-Vocabulary 3D Object Detection via Cycle-Modality Propagation 提出OV-Uni3DETR以解决3D目标检测中的开放词汇问题 open-vocabulary open vocabulary
8 Mitigating Motion Blur in Neural Radiance Fields with Events and Frames 提出新方法以解决神经辐射场中的运动模糊问题 NeRF neural radiance field
9 Mesh2NeRF: Direct Mesh Supervision for Neural Radiance Field Representation and Generation 提出Mesh2NeRF以解决3D生成任务中的辐射场获取问题 NeRF neural radiance field
10 Total-Decom: Decomposed 3D Scene Reconstruction with Minimal Interaction 提出Total-Decom以解决3D场景重建中的对象分解问题 3D reconstruction scene reconstruction ✅
11 Uncertainty-Aware Deep Video Compression with Ensembles 提出不确定性感知深度视频压缩模型以解决压缩效率问题 optical flow motion estimation
12 GlORIE-SLAM: Globally Optimized RGB-only Implicit Encoding Point Cloud SLAM 提出GlORIE-SLAM以解决RGB-only SLAM的全局一致性问题 monocular depth ✅
13 Efficient 3D Instance Mapping and Localization with Neural Fields 提出3DIML框架以解决3D实例分割与定位问题 scene understanding

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#题目一句话要点标签🔗⭐
14 InterDreamer: Zero-Shot Text to 3D Dynamic Human-Object Interaction 提出InterDreamer以解决3D动态人机交互生成问题 world model world models dreamer
15 RSMamba: Remote Sensing Image Classification with State Space Model 提出RSMamba以解决遥感图像分类的复杂性问题 Mamba SSM state space model ✅
16 Neural Fields for 3D Tracking of Anatomy and Surgical Instruments in Monocular Laparoscopic Video Clips 提出神经场方法以实现单目腹腔镜视频中的解剖结构与手术器械的联合跟踪 MAE 3D reconstruction scene reconstruction
17 LocCa: Visual Pretraining with Location-aware Captioners 提出LocCa以解决视觉预训练中位置感知信息不足的问题 visual pre-training
18 De-confounded Data-free Knowledge Distillation for Handling Distribution Shifts 提出因果干预框架以解决数据无关知识蒸馏中的分布偏移问题 distillation
19 GeoAuxNet: Towards Universal 3D Representation Learning for Multi-sensor Point Clouds 提出GeoAuxNet以解决多传感器点云的通用3D表示学习问题 representation learning
20 Siamese Vision Transformers are Scalable Audio-visual Learners 提出音视频共享网络以解决传统方法的可扩展性问题 contrastive learning visual pre-training ✅
21 The Bad Batches: Enhancing Self-Supervised Learning in Image Classification Through Representative Batch Curation 通过代表性批次策划提升自监督学习在图像分类中的表现 representation learning contrastive learning

🔬 支柱九：具身大模型 (Embodied Foundation Models) (5 篇)

#题目一句话要点标签🔗⭐
22 Plug-and-Play Grounding of Reasoning in Multimodal Large Language Models 提出P2G框架以解决多模态大语言模型的推理细节捕捉问题 large language model multimodal instruction following
23 Img2Loc: Revisiting Image Geolocalization using Multi-modality Foundation Models and Image-based Retrieval-Augmented Generation 提出Img2Loc以解决图像地理定位问题 foundation model
24 Towards Multimodal Video Paragraph Captioning Models Robust to Missing Modality 提出MR-VPC框架以解决多模态视频段落描述中的模态缺失问题 multimodal
25 MagicLens: Self-Supervised Image Retrieval with Open-Ended Instructions 提出MagicLens以解决开放式指令下的图像检索问题 foundation model ✅
26 OmniParser: A Unified Framework for Text Spotting, Key Information Extraction and Table Recognition 提出OmniParser以解决视觉文本解析的多任务问题 large language model ✅

🔬 支柱四：生成式动作 (Generative Motion) (3 篇)

#题目一句话要点标签🔗⭐
27 BAMM: Bidirectional Autoregressive Motion Model 提出双向自回归运动模型以解决文本到运动生成的局限性 text-to-motion motion generation motion tokenizer ✅
28 Beyond Talking -- Generating Holistic 3D Human Dyadic Motion for Communication 提出一种生成3D人类对话运动的新方法以增强人际沟通 VQ-VAE human motion
29 RecDiffusion: Rectangling for Image Stitching with Diffusion Models 提出RecDiffusion以解决图像拼接中的非矩形边界问题 motion diffusion model MDM motion diffusion ✅

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
30 OAKINK2: A Dataset of Bimanual Hands-Object Manipulation in Complex Task Completion 提出OAKINK2数据集以解决双手物体操作任务的复杂性问题 manipulation bi-manual bimanual manipulation
31 ShapeFusion: A 3D diffusion model for localized shape editing 提出ShapeFusion以解决3D形状局部编辑问题 manipulation ✅

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
32 RELI11D: A Comprehensive Multimodal Human Motion Dataset and Method 提出RELI11D数据集与LEIR方法以解决复杂人类动作识别问题 human motion multimodal

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

#题目一句话要点标签🔗⭐
33 X-MIC: Cross-Modal Instance Conditioning for Egocentric Action Generalization 提出X-MIC以解决自我中心视频分类中的跨模态适应问题 egocentric Ego4D ✅

⬅️ 返回 cs.CV 首页 · 🏠 返回主页