cs.CV（2025-01-10）

📊 共 19 篇论文 | 🔗 9 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (8 🔗5) 支柱三：空间感知与语义 (Perception & Semantics) (5 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1) 支柱五：交互与反应 (Interaction & Reaction) (1 🔗1)

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

#题目一句话要点标签🔗⭐
1 Valley2: Exploring Multimodal Models with Scalable Vision-Language Design Valley2：探索可扩展视觉-语言设计的多模态模型，提升电商和短视频场景性能。 large language model multimodal ✅
2 Text-to-Edit: Controllable End-to-End Video Ad Creation via Multimodal LLMs 提出Text-to-Edit框架，利用多模态LLM实现可控的端到端视频广告创作 large language model multimodal
3 Towards Iris Presentation Attack Detection with Foundation Models 利用DinoV2和VisualOpenClip等预训练模型提升虹膜活体检测性能 foundation model
4 BRIGHT: A globally distributed multimodal building damage assessment dataset with very-high-resolution for all-weather disaster response 提出BRIGHT多模态建筑损伤评估数据集，用于全天候灾害响应AI模型训练。 multimodal ✅
5 A Multimodal Dataset for Enhancing Industrial Task Monitoring and Engagement Prediction 提出MIAM多模态数据集，用于提升工业任务监控和人机协作中的行为预测。 multimodal ✅
6 PEACE: Empowering Geologic Map Holistic Understanding with MLLMs 提出GeoMap-Agent，赋能多模态大语言模型理解地质图，提升地质调查效率。 large language model multimodal
7 Scalable Vision Language Model Training via High Quality Data Curation SAIL-VL：通过高质量数据构建实现可扩展的视觉语言模型训练 multimodal ✅
8 VideoRAG: Retrieval-Augmented Generation over Video Corpus 提出VideoRAG，通过检索增强生成提升视频语料上的问答准确性 multimodal ✅

🔬 支柱三：空间感知与语义 (Perception & Semantics) (5 篇)

#题目一句话要点标签🔗⭐
9 UV-Attack: Physical-World Adversarial Attacks for Person Detection via Dynamic-NeRF-based UV Mapping 提出UV-Attack，利用动态NeRF的UV映射实现对行人检测器的高成功率物理对抗攻击。 NeRF neural radiance field SMPL ✅
10 StarGen: A Spatiotemporal Autoregression Framework with Video Diffusion Model for Scalable and Controllable Scene Generation StarGen：基于视频扩散模型的时空自回归框架，实现可控可扩展的场景生成 scene reconstruction spatiotemporal ✅
11 Locality-aware Gaussian Compression for Fast and High-quality Rendering 提出LocoGS以解决3D高斯压缩与渲染速度问题 3D gaussian splatting 3DGS gaussian splatting
12 Self-Supervised Partial Cycle-Consistency for Multi-View Matching 提出自监督偏置循环一致性，用于多视角匹配 scene understanding
13 UltraRay: Introducing Full-Path Ray Tracing in Physics-Based Ultrasound Simulation UltraRay：提出基于物理的全路径光线追踪超声模拟方法，提升图像真实感。 scene reconstruction

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

#题目一句话要点标签🔗⭐
14 LlamaV-o1: Rethinking Step-by-step Visual Reasoning in LLMs 提出LlamaV-o1，通过多步课程学习提升LLM在视觉推理任务中的性能 curriculum learning large language model multimodal
15 Overcoming Language Priors for Visual Question Answering Based on Knowledge Distillation 提出KDAR方法，利用知识蒸馏克服视觉问答中的语言先验依赖问题。 distillation multimodal
16 MS-Temba: Multi-Scale Temporal Mamba for Understanding Long Untrimmed Videos 提出MS-Temba，利用多尺度时间Mamba模型解决长未分割视频中的时间动作检测问题。 Mamba SSM
17 From My View to Yours: Ego-to-Exo Transfer in VLMs for Understanding Activities of Daily Living 提出Ego2ExoVLM，解决VLM在日常活动理解中视角不变性导致的自中心属性感知不足问题。 distillation human-object interaction egocentric ✅
18 Deep Reversible Consistency Learning for Cross-modal Retrieval 提出深度可逆一致性学习(DRCL)用于提升跨模态检索的性能与灵活性。 representation learning multimodal

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

#题目一句话要点标签🔗⭐
19 PersonaHOI: Effortlessly Improving Personalized Face with Human-Object Interaction Generation PersonaHOI：融合通用与个性化扩散模型，提升人与物体交互图像生成效果 human-object interaction HOI ✅

⬅️ 返回 cs.CV 首页 · 🏠 返回主页