cs.CV（2024-11-20）

📊 共 30 篇论文 | 🔗 13 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (13 🔗5) 支柱三：空间感知与语义 (Perception & Semantics) (7 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (7 🔗4) 支柱六：视频提取与匹配 (Video Extraction) (2 🔗1) 支柱四：生成式动作 (Generative Motion) (1 🔗1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (13 篇)

#题目一句话要点标签🔗⭐
1 Unsupervised Foundation Model-Agnostic Slide-Level Representation Learning 提出COBRA方法以解决病理全切片图像表示学习问题 Mamba representation learning foundation model ✅
2 FAST-Splat: Fast, Ambiguity-Free Semantics Transfer in Gaussian Splatting FAST-Splat：快速无歧义的高斯溅射语义迁移方法 distillation gaussian splatting splatting
3 XMask3D: Cross-modal Mask Reasoning for Open Vocabulary 3D Semantic Segmentation 提出XMask3D，通过跨模态掩码推理实现开放词汇3D语义分割。 distillation open-vocabulary open vocabulary ✅
4 Efficient Masked AutoEncoder for Video Object Counting and A Large-Scale Benchmark 提出密度嵌入高效掩码自编码计数框架(E-MAC)，解决视频对象计数中前景-背景动态不平衡问题。 representation learning masked autoencoder optical flow
5 Extending Video Masked Autoencoders to 128 frames 提出长视频掩码自编码器(LVMAE)，有效处理128帧视频，提升视频理解性能。 masked autoencoder MAE foundation model
6 MambaDETR: Query-based Temporal Modeling using State Space Model for Multi-View 3D Object Detection MambaDETR：利用状态空间模型进行多视角3D目标检测的查询式时序建模 Mamba state space model
7 Decompose and Leverage Preferences from Expert Models for Improving Trustworthiness of MLLMs 提出DecompGen框架，利用专家模型分解评估MLLM响应，提升其可信度。 preference learning large language model multimodal
8 Intensity-Spatial Dual Masked Autoencoder for Multi-Scale Feature Learning in Chest CT Segmentation 提出强度-空间双掩码自编码器(ISD-MAE)用于胸部CT多尺度特征学习与分割 masked autoencoder MAE contrastive learning ✅
9 Find Any Part in 3D 利用2D基础模型驱动的数据引擎，实现任意3D物体部件的开放世界分割 world model foundation model ✅
10 Identity Preserving 3D Head Stylization with Multiview Score Distillation 提出基于多视角Score Distillation的3D头部风格化方法，提升身份保持能力 distillation ✅
11 Cross-Camera Distracted Driver Classification through Feature Disentanglement and Contrastive Learning 提出DBMNet，通过特征解耦和对比学习实现跨摄像头分心驾驶员分类。 contrastive learning
12 Collaborative Feature-Logits Contrastive Learning for Open-Set Semi-Supervised Object Detection 提出CFL-Detector，解决开放集半监督目标检测中的OOD误分类问题 contrastive learning
13 RobustFormer: Noise-Robust Pre-training for images and videos RobustFormer：一种噪声鲁棒的图像和视频预训练方法，利用DWT提升Transformer在噪声环境下的性能。 masked autoencoder MAE

🔬 支柱三：空间感知与语义 (Perception & Semantics) (7 篇)

#题目一句话要点标签🔗⭐
14 GazeGaussian: High-Fidelity Gaze Redirection with 3D Gaussian Splatting GazeGaussian：基于3D高斯溅射的高保真视线重定向 3D gaussian splatting 3DGS gaussian splatting ✅
15 Generating 3D-Consistent Videos from Unposed Internet Photos 提出一种自监督方法，从无位姿互联网照片生成3D一致性视频 3D gaussian splatting gaussian splatting splatting
16 Sparse Input View Synthesis: 3D Representations and Reliable Priors 针对稀疏视角的新视角合成，提出基于3D表示和可靠先验的解决方案 NeRF neural radiance field optical flow
17 Robust SG-NeRF: Robust Scene Graph Aided Neural Surface Reconstruction 提出鲁棒的SG-NeRF，利用场景图辅助神经表面重建，解决相机姿态噪声问题。 NeRF
18 DATAP-SfM: Dynamic-Aware Tracking Any Point for Robust Structure from Motion in the Wild DATAP-SfM：动态感知追踪任意点，实现野外场景鲁棒的运动结构重建 depth estimation optical flow
19 Geometric Algebra Planes: Convex Implicit Neural Volumes 提出GA-Planes：一种可凸优化训练的隐式神经场表示方法，用于体积建模。 implicit representation
20 Practical Compact Deep Compressed Sensing 提出PCNet，一种实用紧凑的深度压缩感知网络，提升图像重建质量。 implicit representation ✅

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

#题目一句话要点标签🔗⭐
21 VideoAutoArena: An Automated Arena for Evaluating Large Multimodal Models in Video Analysis through User Simulation VideoAutoArena：通过用户模拟自动评估视频分析大模型的竞技场基准 multimodal
22 Adapting Vision Foundation Models for Robust Cloud Segmentation in Remote Sensing Images 提出Cloud-Adapter，利用视觉基础模型实现鲁棒的遥感图像云分割 foundation model ✅
23 Hints of Prompt: Enhancing Visual Representation for Multimodal LLMs in Autonomous Driving 提出Hints of Prompt (HoP)框架，增强多模态LLM在自动驾驶场景中的视觉表征能力 multimodal
24 MEGL: Multimodal Explanation-Guided Learning 提出MEGL：一种多模态解释引导学习框架，提升图像分类模型的可解释性和性能。 multimodal
25 Unsupervised Homography Estimation on Multimodal Image Pair via Alternating Optimization 提出AltO，通过交替优化解决多模态图像对的无监督单应性估计问题 multimodal ✅
26 On the Consistency of Video Large Language Models in Temporal Comprehension 针对视频大语言模型时间理解一致性问题，提出事件时序验证调优方法 large language model ✅
27 FabuLight-ASD: Unveiling Speech Activity via Body Language FabuLight-ASD：利用身体语言增强多模态环境下的语音活动检测 multimodal ✅

🔬 支柱六：视频提取与匹配 (Video Extraction) (2 篇)

#题目一句话要点标签🔗⭐
28 DIS-Mine: Instance Segmentation for Disaster-Awareness in Poor-Light Condition in Underground Mines DIS-Mine：针对地下矿井弱光环境的灾害感知实例分割方法 feature matching
29 X as Supervision: Contending with Depth Ambiguity in Unsupervised Monocular 3D Pose Estimation 提出基于多假设检测与3D先验的无监督单目3D姿态估计方法 SMPL ✅

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
30 REDUCIO! Generating 1K Video within 16 Seconds using Extremely Compressed Motion Latents REDUCIO：利用极度压缩的运动潜在空间，在16秒内生成1K视频 motion latent ✅

⬅️ 返回 cs.CV 首页 · 🏠 返回主页