cs.CV（2024-03-14）

📊 共 52 篇论文 | 🔗 24 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (21 🔗9) 支柱二：RL算法与架构 (RL & Architecture) (12 🔗7) 支柱三：空间感知与语义 (Perception & Semantics) (11 🔗6) 支柱六：视频提取与匹配 (Video Extraction) (4 🔗1) 支柱七：动作重定向 (Motion Retargeting) (2) 支柱一：机器人控制 (Robot Control) (1 🔗1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (21 篇)

#题目一句话要点标签🔗⭐
1 Images are Achilles' Heel of Alignment: Exploiting Visual Vulnerabilities for Jailbreaking Multimodal Large Language Models 提出HADES方法以解决多模态大语言模型的对齐脆弱性问题 large language model multimodal ✅
2 UniCode: Learning a Unified Codebook for Multimodal Large Language Models 提出UniCode以解决多模态大语言模型的编码限制问题 large language model multimodal
3 MM1: Methods, Analysis & Insights from Multimodal LLM Pre-training 提出MM1以优化多模态大语言模型的预训练方法 large language model multimodal chain-of-thought
4 VisionGPT-3D: A Generalized Multimodal Agent for Enhanced 3D Vision Understanding 提出VisionGPT-3D以解决3D视觉理解中的多模态融合问题 large language model foundation model multimodal
5 VisionGPT: Vision-Language Understanding Agent Using Generalized Multimodal Framework 提出VisionGPT以解决开放世界视觉感知问题 large language model foundation model multimodal
6 Eyes Closed, Safety On: Protecting Multimodal LLMs via Image-to-Text Transformation 提出ECSO以增强多模态大语言模型的安全性 large language model multimodal
7 Griffon v2: Advancing Multimodal Perception with High-Resolution Scaling and Visual-Language Co-Referring 提出Griffon v2以解决多模态感知中的高分辨率限制问题 large language model multimodal ✅
8 Towards Comprehensive Multimodal Perception: Introducing the Touch-Language-Vision Dataset 提出TLV数据集以解决多模态感知中的语言缺失问题 multimodal ✅
9 OneTracker: Unifying Visual Object Tracking with Foundation Models and Efficient Tuning 提出OneTracker以统一视觉目标跟踪任务 foundation model
10 M&M: Multimodal-Multitask Model Integrating Audiovisual Cues in Cognitive Load Assessment 提出M&M模型以解决认知负荷评估中的多模态融合问题 multimodal
11 E2E-MFD: Towards End-to-End Synchronous Multimodal Fusion Detection 提出E2E-MFD以解决多模态图像融合与目标检测的复杂训练问题 multimodal ✅
12 Annotation Free Semantic Segmentation with Vision Foundation Models 提出无注释语义分割方法以解决数据标注成本高的问题 foundation model
13 Task-Specific Adaptation of Segmentation Foundation Model via Prompt Learning 提出任务特定适应方法以解决SAM实例分割问题 foundation model
14 Leveraging Foundation Model Automatic Data Augmentation Strategies and Skeletal Points for Hands Action Recognition in Industrial Assembly Lines 提出基于基础模型的数据增强策略以解决工业装配线手部动作识别问题 foundation model
15 XCoOp: Explainable Prompt Learning for Computer-Aided Diagnosis via Concept-guided Context Optimization 提出可解释的提示学习框架以解决医疗诊断中的解释性问题 large language model foundation model
16 GiT: Towards Generalist Vision Transformer through Universal Language Interface 提出GiT框架以解决视觉任务模块化不足问题 large language model foundation model ✅
17 Explore In-Context Segmentation via Latent Diffusion Models 提出潜在扩散模型以解决上下文分割问题 foundation model
18 Introducing Routing Functions to Vision-Language Parameter-Efficient Fine-Tuning with Low-Rank Bottlenecks 提出路由函数以增强视觉-语言参数高效微调 multimodal ✅
19 B-AVIBench: Towards Evaluating the Robustness of Large Vision-Language Model on Black-box Adversarial Visual-Instructions 提出B-AVIBench以评估大型视觉语言模型的鲁棒性 multimodal ✅
20 Can We Talk Models Into Seeing the World Differently? 研究多模态模型中的视觉偏差与语言提示的影响 large language model ✅
21 PYRA: Parallel Yielding Re-Activation for Training-Inference Efficient Task Adaptation 提出PYRA以解决大规模模型训练与推理效率问题 foundation model ✅

🔬 支柱二：RL算法与架构 (RL & Architecture) (12 篇)

#题目一句话要点标签🔗⭐
22 3D-VLA: A 3D Vision-Language-Action Generative World Model 提出3D-VLA以解决现有VLA模型缺乏3D世界整合的问题 world model world models vision-language-action
23 Dyadic Interaction Modeling for Social Behavior Generation 提出双人互动建模以生成社交行为 contrastive learning motion generation VQ-VAE ✅
24 Open-Vocabulary Object Detection with Meta Prompt Representation and Instance Contrastive Optimization 提出Meta Prompt与实例对比优化以解决开放词汇目标检测问题 contrastive learning distillation open-vocabulary
25 FastSAM3D: An Efficient Segment Anything Model for 3D Volumetric Medical Images 提出FastSAM3D以解决3D医学图像分割的高延迟问题 distillation sam 3D SAM 3D ✅
26 MambaTalk: Efficient Holistic Gesture Synthesis with Selective State Space Models 提出MambaTalk以解决高延迟手势合成问题 Mamba SSM state space model ✅
27 SELECTOR: Heterogeneous graph network with convolutional masked autoencoder for multimodal robust prediction of cancer survival 提出SELECTOR以解决癌症患者生存率预测中的多模态数据缺失问题 masked autoencoder multimodal ✅
28 Video Mamba Suite: State Space Model as a Versatile Alternative for Video Understanding 提出视频Mamba套件以提升视频理解能力 Mamba state space model ✅
29 LocalMamba: Visual State Space Model with Windowed Selective Scan 提出LocalMamba以优化视觉状态空间模型的扫描策略 Mamba state space model ✅
30 GroupContrast: Semantic-aware Self-supervised Representation Learning for 3D Understanding 提出GroupContrast以解决3D表示学习中的语义冲突问题 representation learning contrastive learning scene understanding
31 VM-UNET-V2 Rethinking Vision Mamba UNet for Medical Image Segmentation 提出VM-UNET-V2以解决医学图像分割中的长距离依赖问题 Mamba SSM state space model ✅
32 3D-SceneDreamer: Text-Driven 3D-Consistent Scene Generation 提出3D-SceneDreamer以解决3D场景生成中的一致性问题 dreamer NeRF
33 Unsupervised Modality-Transferable Video Highlight Detection with Representation Activation Sequence Learning 提出无监督模态可转移的视频亮点检测方法以解决标注困难问题 contrastive learning multimodal

🔬 支柱三：空间感知与语义 (Perception & Semantics) (11 篇)

#题目一句话要点标签🔗⭐
34 Relaxing Accurate Initialization Constraint for 3D Gaussian Splatting 提出RAIN-GS以解决3D Gaussian Splatting对初始化的依赖问题 3D gaussian splatting 3DGS 3D reconstruction ✅
35 PosSAM: Panoptic Open-vocabulary Segment Anything 提出PosSAM以解决开放词汇全景分割问题 open-vocabulary open vocabulary ✅
36 Renovating Names in Open-Vocabulary Segmentation Benchmarks 提出RENOVATE框架以提升开放词汇分割基准的名称质量 open-vocabulary open vocabulary
37 Intention-driven Ego-to-Exo Video Generation 提出意图驱动的自我到外部视频生成框架以解决视角转换问题 optical flow egocentric spatiotemporal
38 The NeRFect Match: Exploring NeRF Features for Visual Localization 提出NeRFMatch以解决视觉定位中的2D-3D匹配问题 NeRF neural radiance field
39 PreSight: Enhancing Autonomous Vehicle Perception with City-Scale NeRF Priors 提出PreSight以解决自动驾驶感知系统在复杂环境中的不足 NeRF neural radiance field ✅
40 SD-Net: Symmetric-Aware Keypoint Prediction and Domain Adaptation for 6D Pose Estimation In Bin-picking Scenarios 提出SD-Net以解决对称物体在6D姿态估计中的挑战 6D pose estimation ✅
41 Hyper-3DG: Text-to-3D Gaussian Generation via Hypergraph 提出Hyper-3DG以解决3D生成中的高阶关联问题 3DGS ✅
42 MARVIS: Motion & Geometry Aware Real and Virtual Image Segmentation 提出MARVIS以解决水面图像分割中的真实与虚拟区域识别问题 3D reconstruction geometric consistency ✅
43 RoDUS: Robust Decomposition of Static and Dynamic Elements in Urban Scenes 提出RoDUS以解决城市场景中静态与动态元素分离问题 NeRF
44 Improving Distant 3D Object Detection Using 2D Box Supervision 提出LR3D框架以解决远距离3D物体检测问题 depth estimation

🔬 支柱六：视频提取与匹配 (Video Extraction) (4 篇)

#题目一句话要点标签🔗⭐
45 Distribution and Depth-Aware Transformers for 3D Human Mesh Recovery 提出D2A-HMR以解决3D人类网格恢复中的深度模糊问题 human mesh recovery HMR
46 ThermoHands: A Benchmark for 3D Hand Pose Estimation from Egocentric Thermal Images 提出ThermoHands以解决复杂环境下3D手势估计问题 egocentric
47 On the Utility of 3D Hand Poses for Action Recognition 提出HandFormer以解决3D手势在动作识别中的不足 egocentric multimodal
48 Score-Guided Diffusion for 3D Human Recovery 提出ScoreHMR以解决3D人类姿态与形状重建问题 human mesh recovery ✅

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

#题目一句话要点标签🔗⭐
49 GazeMotion: Gaze-guided Human Motion Forecasting 提出GazeMotion以解决人类动作预测中的视线引导问题 human motion
50 Adversarial Training with OCR Modality Perturbation for Scene-Text Visual Question Answering 提出多模态对抗训练以解决场景文本视觉问答中的OCR依赖问题 spatial relationship multimodal

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
51 OpenGraph: Open-Vocabulary Hierarchical 3D Graph Representation in Large-Scale Outdoor Environments 提出OpenGraph以解决大规模户外环境的开放词汇图表示问题 manipulation open-vocabulary open vocabulary ✅

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

#题目一句话要点标签🔗⭐
52 rFaceNet: An End-to-End Network for Enhanced Physiological Signal Extraction through Identity-Specific Facial Contours 提出rFaceNet以增强生理信号提取的准确性 PULSE

⬅️ 返回 cs.CV 首页 · 🏠 返回主页