cs.CV（2024-04-18）

📊 共 28 篇论文 | 🔗 10 篇有代码

🎯 兴趣领域导航

支柱三：空间感知与语义 (Perception & Semantics) (9 🔗4) 支柱九：具身大模型 (Embodied Foundation Models) (7 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗2) 支柱四：生成式动作 (Generative Motion) (2) 支柱一：机器人控制 (Robot Control) (2) 支柱七：动作重定向 (Motion Retargeting) (2 🔗1) 支柱八：物理动画 (Physics-based Animation) (1 🔗1)

🔬 支柱三：空间感知与语义 (Perception & Semantics) (9 篇)

#题目一句话要点标签🔗⭐
1 BLINK: Multimodal Large Language Models Can See but Not Perceive 提出BLINK基准以解决多模态语言模型的视觉感知问题 depth estimation large language model multimodal
2 SPIdepth: Strengthened Pose Information for Self-supervised Monocular Depth Estimation 提出SPIdepth以增强姿态信息解决单目深度估计问题 depth estimation monocular depth scene understanding ✅
3 The devil is in the object boundary: towards annotation-free instance segmentation using Foundation Models 提出Zip以解决实例分割中的注释依赖问题 open-vocabulary open vocabulary foundation model ✅
4 AG-NeRF: Attention-guided Neural Radiance Fields for Multi-height Large-scale Outdoor Scene Rendering 提出AG-NeRF以解决多高度大规模户外场景渲染问题 NeRF neural radiance field
5 6Img-to-3D: Few-Image Large-Scale Outdoor Driving Scene Reconstruction 提出6Img-to-3D以解决少图像大规模户外场景重建问题 3D reconstruction scene reconstruction ✅
6 Dynamic Gaussians Mesh: Consistent Mesh Reconstruction from Dynamic Scenes 提出动态高斯网格以解决动态场景中的网格重建问题 3D gaussian splatting gaussian splatting splatting
7 Moving Object Segmentation: All You Need Is SAM (and Flow) 提出基于SAM和光流的运动目标分割方法 optical flow
8 MeshLRM: Large Reconstruction Model for High-Quality Meshes 提出MeshLRM以实现高质量网格的快速重建 NeRF ✅
9 What does CLIP know about peeling a banana? 提出AffordanceCLIP以解决工具识别与动作关联问题 affordance

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

#题目一句话要点标签🔗⭐
10 From Image to Video, what do we need in multimodal LLMs? 提出RED-VILLM以解决视频LLM结构冗长和数据需求高的问题 large language model foundation model multimodal
11 MedThink: Explaining Medical Visual Question Answering via Multimodal Decision-Making Rationale 提出MedThink以解决医学视觉问答模型可解释性问题 large language model multimodal
12 How to Benchmark Vision Foundation Models for Semantic Segmentation? 提出标准化基准以评估视觉基础模型在语义分割中的表现 foundation model ✅
13 StyleBooth: Image Style Editing with Multimodal Instruction 提出StyleBooth以解决多模态图像风格编辑问题 multimodal ✅
14 V2Xum-LLM: Cross-Modal Video Summarization with Temporal Prompt Instruction Tuning 提出V2Xum-LLM以解决多模态视频摘要数据不足问题 large language model multimodal
15 DeepLocalization: Using change point detection for Temporal Action Localization 提出DeepLocalization以解决驾驶行为监测中的分心驾驶问题 large language model
16 PureForest: A Large-Scale Aerial Lidar and Aerial Imagery Dataset for Tree Species Classification in Monospecific Forests 提出PureForest数据集以解决树种分类的挑战 multimodal

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

#题目一句话要点标签🔗⭐
17 Evaluating Alternatives to SFM Point Cloud Initialization for Gaussian Splatting 提出新初始化策略以解决Gaussian Splatting对SFM依赖的问题 distillation 3D gaussian splatting 3DGS ✅
18 MambaPupil: Bidirectional Selective Recurrent model for Event-based Eye tracking 提出MambaPupil以解决事件驱动眼动追踪中的不稳定性问题 Mamba SSM
19 Not All Voxels Are Equal: Hardness-Aware Semantic Scene Completion with Self-Distillation 提出HASSC以解决语义场景完成中的困难体素问题 distillation ✅
20 Data-free Knowledge Distillation for Fine-grained Visual Categorization 提出DFKD-FGVC以解决细粒度视觉分类中的数据蒸馏问题 distillation
21 Curriculum Point Prompting for Weakly-Supervised Referring Image Segmentation 提出课程点提示以解决弱监督图像分割问题 curriculum learning foundation model

🔬 支柱四：生成式动作 (Generative Motion) (2 篇)

#题目一句话要点标签🔗⭐
22 MultiPhys: Multi-Person Physics-aware 3D Motion Estimation 提出MultiPhys以解决多人物体运动估计问题 foot skating penetration motion estimation
23 MIDGET: Music Conditioned 3D Dance Generation 提出MIDGET以解决音乐与舞蹈生成匹配问题 VQ-VAE

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
24 DoughNet: A Visual Predictive Model for Topological Manipulation of Deformable Objects 提出DoughNet以解决可变形物体的拓扑操控问题 manipulation predictive model
25 TrACT: A Training Dynamics Aware Contrastive Learning Framework for Long-tail Trajectory Prediction 提出TrACT框架以解决长尾轨迹预测问题 motion planning contrastive learning

🔬 支柱七：动作重定向 (Motion Retargeting) (2 篇)

#题目一句话要点标签🔗⭐
26 Partial-to-Partial Shape Matching with Geometric Consistency 提出几何一致性方法以解决部分形状匹配问题 geometric consistency
27 Aligning Actions and Walking to LLM-Generated Textual Descriptions 利用LLM生成文本描述以对齐动作与步态分析 motion representation large language model ✅

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

#题目一句话要点标签🔗⭐
28 Seeing Motion at Nighttime with an Event Camera 提出夜间动态成像方法以解决传统相机的模糊问题 spatiotemporal TAMP ✅

⬅️ 返回 cs.CV 首页 · 🏠 返回主页