cs.CV（2024-04-19）

📊 共 25 篇论文 | 🔗 4 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (8 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (6) 支柱二：RL算法与架构 (RL & Architecture) (4 🔗1) 支柱四：生成式动作 (Generative Motion) (2) 支柱一：机器人控制 (Robot Control) (1 🔗1) 支柱五：交互与反应 (Interaction & Reaction) (1) 支柱七：动作重定向 (Motion Retargeting) (1 🔗1) 支柱八：物理动画 (Physics-based Animation) (1) 支柱六：视频提取与匹配 (Video Extraction) (1)

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

#题目一句话要点标签🔗⭐
1 Groma: Localized Visual Tokenization for Grounding Multimodal Large Language Models 提出Groma以解决多模态大语言模型的视觉定位问题 large language model multimodal visual grounding ✅
2 MoVA: Adapting Mixture of Vision Experts to Multimodal Context 提出MoVA以解决多模态视觉理解的偏差问题 large language model multimodal
3 PDF-MVQA: A Dataset for Multimodal Information Retrieval in PDF-based Visual Question Answering 提出PDF-MVQA以解决文本主导文档的多模态信息检索问题 multimodal
4 Transformer-Based Classification Outcome Prediction for Multimodal Stroke Treatment 提出基于Transformer的多模态融合框架以预测中风治疗效果 multimodal
5 ELEV-VISION-SAM: Integrated Vision Language and Foundation Model for Automated Estimation of Building Lowest Floor Elevation 提出ELEV-VISION-SAM以解决建筑最低楼层高度估计问题 foundation model
6 Data Alignment for Zero-Shot Concept Generation in Dermatology AI 提出数据对齐方法以解决皮肤科AI中的零-shot概念生成问题 large language model foundation model
7 TextSquare: Scaling up Text-Centric Visual Instruction Tuning 提出TextSquare以提升文本中心视觉问答的性能 large language model multimodal
8 Pre-trained Vision-Language Models Learn Discoverable Visual Concepts 提出概念发现与学习框架以提升视觉语言模型的概念理解能力 multimodal

🔬 支柱三：空间感知与语义 (Perception & Semantics) (6 篇)

#题目一句话要点标签🔗⭐
9 Unified Scene Representation and Reconstruction for 3D Large Language Models 提出Uni3DR^2以解决3D场景理解不足的问题 3D reconstruction scene understanding large language model
10 EfficientGS: Streamlining Gaussian Splatting for Large-Scale High-Resolution Scene Representation 提出EfficientGS以解决大规模高分辨率场景表示问题 3D gaussian splatting 3DGS gaussian splatting
11 Learn2Talk: 3D Talking Face Learns from 2D Talking Face 提出Learn2Talk以提升3D人脸动画的唇同步精度 3D gaussian splatting gaussian splatting splatting
12 Weakly Supervised LiDAR Semantic Segmentation via Scatter Image Annotation 提出基于散射图像标注的弱监督LiDAR语义分割方法 optical flow multimodal
13 ECOR: Explainable CLIP for Object Recognition 提出ECOR以解决VLM可解释性不足问题 open-vocabulary open vocabulary
14 3D Multi-frame Fusion for Video Stabilization 提出RStab框架以解决视频稳定性问题 optical flow

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

#题目一句话要点标签🔗⭐
15 Purposer: Putting Human Motion Generation in Context 提出Purposer以解决人类运动生成的上下文问题 representation learning motion generation human motion
16 Look Before You Decide: Prompting Active Deduction of MLLMs for Assumptive Reasoning 提出主动推理方法以提升多模态大语言模型的假设推理能力 reinforcement learning large language model multimodal
17 MambaMOS: LiDAR-based 3D Moving Object Segmentation with Motion-aware State Space Model 提出MambaMOS以解决LiDAR点云中动态物体分割问题 Mamba state space model ✅
18 Zero-Shot Medical Phrase Grounding with Off-the-shelf Diffusion Models 提出基于扩散模型的零-shot医学短语定位方法 contrastive learning VIO foundation model

🔬 支柱四：生成式动作 (Generative Motion) (2 篇)

#题目一句话要点标签🔗⭐
19 MCM: Multi-condition Motion Synthesis Framework 提出MCM框架以解决多条件人类动作合成问题 motion synthesis human motion human motion synthesis
20 Analysis of Classifier-Free Guidance Weight Schedulers 提出分类器无关引导权重调度器以提升图像生成质量 classifier-free guidance

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
21 PhysDreamer: Physics-Based Interaction with 3D Objects via Video Generation 提出PhysDreamer以解决3D物体动态交互问题 manipulation dreamer physically plausible ✅

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

#题目一句话要点标签🔗⭐
22 Exploring Interactive Semantic Alignment for Efficient HOI Detection with Vision-language Model 提出ISA-HOI以解决HOI检测中的上下文信息缺失问题 human-object interaction HOI

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

#题目一句话要点标签🔗⭐
23 Motion-adaptive Separable Collaborative Filters for Blind Motion Deblurring 提出运动自适应可分离协同滤波器以解决盲运动去模糊问题 motion estimation ✅

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

#题目一句话要点标签🔗⭐
24 On-Demand Earth System Data Cubes 提出cubo工具以高效生成AI专用地球系统数据立方体 spatiotemporal

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

#题目一句话要点标签🔗⭐
25 SkelFormer: Markerless 3D Pose and Shape Estimation using Skeletal Transformers 提出SkelFormer以解决无标记3D姿态与形状估计问题 markerless motion capture

⬅️ 返回 cs.CV 首页 · 🏠 返回主页