cs.CV（2026-09-23）

📊 共 26 篇论文 | 🔗 8 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (11 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (7 🔗3) 支柱三：空间感知与语义 (Perception & Semantics) (4 🔗1) 支柱一：机器人控制 (Robot Control) (3 🔗1) 支柱七：动作重定向 (Motion Retargeting) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (11 篇)

#题目一句话要点标签🔗⭐
1 AnchorReasoning: A Visual Grounding and Causal Reasoning Dataset in Long-Tail Autonomous Driving Scenarios 提出AnchorReasoning以解决长尾自主驾驶中的视觉推理问题 embodied AI visual grounding chain-of-thought
2 AstraLOD3: Zero-shot multimodal agentic reconstruction of LOD3 building models 提出AstraLOD3以解决LOD3建筑模型重建的灵活性问题 foundation model multimodal
3 CereVLA: Cerebellum-Inspired Consequence-Aware Residual Governance for Efficient Vision-Language-Action Execution 提出CereVLA以解决视觉-语言-行动执行中的反馈不足问题 vision-language-action VLA
4 Benchmarking Hyperspectral Foundation Models for Hyperspectral Unmixing 基于基础模型的高光谱解混合性能评估与特征提升方法 foundation model
5 A Unified Framework and Dataset for Oriented Object Visual Grounding in Remote Sensing 提出O$^2$-VG框架以解决遥感图像中定向物体视觉定位问题 visual grounding ✅
6 FFM-CP: Cross-Backbone Fusion of Vision-Language Foundation Models for Few-Shot Computational Pathology 提出FFM-CP以解决病理图像少样本学习问题 foundation model
7 A generalizable structural brain MRI foundation model built through dual-priority federated pretraining 提出BrainFedFM以解决结构性脑MRI分析中的隐私与数据分散问题 foundation model
8 Invisible in Space, Visible in Time: Motion Vision CAPTCHA against GUI Agents 提出运动视觉CAPTCHA以应对GUI代理的挑战 large language model multimodal
9 Privacy-Preserving Semantic Segmentation from High-Resolution Depth and Ultra-Low-Resolution RGB 提出高分辨率深度与超低分辨率RGB结合的隐私保护语义分割方法 zero-shot transfer
10 EmbodiedMemory-Bench: Benchmarking Embodied Memory for Long-Horizon Embodied Tasks 提出EmbodiedMemory-Bench以解决长时间体交互中的记忆问题 multimodal ✅
11 All modalities are equal, but video is more equal: Closing the Cross-Attention Gap in Joint Video Generation 提出RecCAR以解决视频生成中的跨模态对应差距问题 multimodal

🔬 支柱二：RL算法与架构 (RL & Architecture) (7 篇)

#题目一句话要点标签🔗⭐
12 Hybrid Gaussians for Robust Open-Vocabulary 3D Segmentation with Multi-View Object Association and Boundary Refinement 提出混合高斯模型以解决开放词汇3D分割问题 contrastive learning open-vocabulary open vocabulary ✅
13 Do Center Biases Propagate? Robustness of Pathology Foundation Models in Whole-Slide Image Classification 评估病理基础模型在全切片图像分类中的中心偏差传播问题 representation learning foundation model
14 Depth-Guided Contrastive Learning for 2D Representations with 3D Spatial Awareness 提出深度引导对比学习以增强2D表示的3D空间感知 representation learning contrastive learning scene understanding ✅
15 Geometry-Conditioned Visual Place Recognition in Natural Environments 提出深度感知蒸馏方法以解决自然环境中的视觉位置识别问题 distillation geometric consistency foundation model
16 PBLH Estimation from Satellite Radiances via a Dual-Encoder Transformer 提出双编码器Transformer以解决卫星辐射数据的PBLH估计问题 MAE multimodal
17 DeltaS: Reading the Gated Linear Attention State for KV Cache Eviction in Streaming Video 提出DeltaS以解决长视频流中的KV缓存驱逐问题 linear attention ✅
18 Prompt, Probe, Train, or Annotate? Single-camera sports video understanding in amateur settings 提出多种方法以解决业余体育视频理解问题 world model world models

🔬 支柱三：空间感知与语义 (Perception & Semantics) (4 篇)

#题目一句话要点标签🔗⭐
19 GaussPDE: Graph-Based Partial Differential Equation-Driven Rendering for 3D Gaussian Splatting 提出GaussPDE以解决3D高斯场景动态渲染问题 3D gaussian splatting 3DGS gaussian splatting
20 Know-Your-Scene (KYS)-SLAM: Hierarchical Semantic-Motion Priors for Feature Matching in Stereo Visual SLAM 提出KYS-SLAM以解决立体视觉SLAM中的特征匹配问题 visual SLAM optical flow feature matching
21 MotionSpec: Spectral Trajectory Supervision for Motion-Consistent Video Generation 提出MotionSpec以解决视频生成中的运动一致性问题 optical flow
22 Task-Induced Riemannian Metrics for Vision Transformer Feature Spaces 提出任务诱导的黎曼度量以优化视觉变换器特征空间 VGGT ✅

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
23 Latent evolving World Action Model 提出LeWAM以提高世界动作模型的效率与可扩展性 manipulation imitation learning DPO
24 Frozen Flows Forget: Diagnosing and Restoring Lost Motion in a Latent-flow World Model 提出DART方法以解决潜在流模型中的运动丢失问题 manipulation world model world models
25 SatUnreal: A High-Precision Synthetic Dataset for Satellite Stereo Matching via Unreal Engine 提出SatUnreal以解决卫星立体匹配数据集不足问题 sim-to-real 3D reconstruction zero-shot transfer ✅

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

#题目一句话要点标签🔗⭐
26 ZoomDiff: A High-Fidelity Diffusion Model for Dual-Camera Smooth Zooming 提出ZoomDiff以解决双摄像头平滑变焦问题 geometric consistency ✅

⬅️ 返回 cs.CV 首页 · 🏠 返回主页