cs.CV（2026-10-05）

📊 共 38 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) (17 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (9 🔗5) 支柱三：空间感知与语义 (Perception & Semantics) (6) 支柱一：机器人控制 (Robot Control) (3) 支柱八：物理动画 (Physics-based Animation) (2) 支柱五：交互与反应 (Interaction & Reaction) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (17 篇)

#题目一句话要点标签🔗⭐
1 Anatomy-aware Fine-grained Multimodal Fusion for Laryngopharyngeal Cancer T-Staging Prediction Using CT and Radiology Report 提出解剖结构感知的多模态融合方法以解决喉咽癌T分期预测问题 multimodal
2 BrainTRACE: Tracing Longitudinal, Multimodal, and Volumetric Evidence in Brain MRI Clinical Reasoning 提出BrainTRACE以解决脑MRI临床推理问题 multimodal
3 MTOR: Generalizable AI-Generated Video Detection with Multimodal Semantics and Temporal Over-Regularity 提出MTOR以解决AI生成视频检测的挑战 multimodal ✅
4 Label-Free Coreset Selection with Foundation Models for Efficient Annotation in Computational Pathology 提出GCcore以解决计算病理学中标注数据不足问题 foundation model ✅
5 FairRSFM: A Biome-Aware Benchmark and Debiasing Framework for Remote Sensing Foundation Models 提出FairRSFM以解决遥感基础模型的生态区域评估问题 foundation model ✅
6 StageVLN: Spatial and Trajectory Auxiliary Guidance for Efficient Vision-Language Navigation 提出StageVLN以解决视觉语言导航中的几何信息缺失问题 VLN foundation model
7 Learning to Read the Contextual Tokens in Diffusion Transformers 提出一种框架以解读扩散变换器中的上下文标记 large language model multimodal
8 Environmental sensor readings in two crop disease image datasets identify the session in which each image was taken 揭示环境传感器数据对作物病害分类的影响 multimodal TAMP
9 From Transformation to Target State: Rethinking Query Representation for Zero-Shot Composed Image Retrieval 提出ASAP-CIR框架以解决零-shot复合图像检索问题 large language model multimodal
10 VideoTapestry: Query-Adaptive Memory Refinement for Multi-Agent Long-Video Understanding 提出VideoTapestry以解决长视频理解中的记忆适应性问题 multimodal
11 Readout Blindness: VLM Scores Miss the Spatial Direction Their Frozen Encoders Retain 提出抗对称位移读出方法以解决视觉语言模型的方向性盲点问题 multimodal
12 MoCAR: Motion-code Coordinate-aware AutoRegression for Continuous Trajectory Forecasting 提出MoCAR以解决连续轨迹预测中的多模态挑战 multimodal
13 Efficient Test-time Adaptation through Candidate Verification and Divergence Shifts 提出测试时候适应的新方法以解决目标领域偏移问题 zero-shot transfer
14 ROT: Rotating Hidden States towards Contextual Vectors for Hallucination Mitigation in LVLMs 提出ROT框架以缓解LVLM中的对象幻觉问题 multimodal
15 UltraDub: Towards Authentic Dubbing by Unifying Visually-Steered Flow Learning and Trajectory Guidance 提出UltraDub以解决视觉语音克隆中的同步与一致性问题 multimodal
16 Prompt and Refinement: Asymmetric Mutual Learning for Infrared Small Target Detection with Noisy Labels 提出PAR方法以解决红外小目标检测中的标签噪声问题 foundation model
17 Fitting Vision Adapters at Frontier Scales 提出视觉适配器以增强语言模型的视觉能力 multimodal

🔬 支柱二：RL算法与架构 (RL & Architecture) (9 篇)

#题目一句话要点标签🔗⭐
18 T-JEPA: A Temporal Joint-Embedding Predictive Architecture for Learning Better Remote Sensing Representations 提出T-JEPA以解决遥感数据时间序列建模问题 JEPA Joint-Embedding Predictive Architecture joint-embedding predictive architecture
19 VepAgent: Bridging Causal-Transition via Tool-Augmented Reinforcement Learning for Video Event Prediction 提出VepAgent以解决视频事件预测中的因果转变问题 reinforcement learning large language model multimodal
20 fMRI-TAMCL: Text-Anchored Supervised Multimodal Contrastive Learning for fMRI-Based Brain Disorder Classification 提出fMRI-TAMCL以解决fMRI脑部疾病分类中的多模态融合问题 contrastive learning multimodal
21 HLA-WM: Hybrid Linear Attention for Long-Horizon Video World Models 提出HLA-WM以解决长视频世界模型中的长程遗忘问题 world model world models linear attention ✅
22 RealtimeWAM: One-Step Asynchronous World Action Models 提出RealtimeWAM以解决视频行动预测效率低下问题 world action model world action models distillation ✅
23 Imagine to Act: High-Fidelity Data Synthesis via Image Editing World Model for Scalable GUI Agent Training 提出Infinite-Dreamer以解决高保真数据合成问题 world model world models dreamer ✅
24 MeSD: Multi-Evidence Self-Distillation for VideoLLM 提出MeSD框架以解决视频大语言模型中的自蒸馏问题 reinforcement learning distillation privileged information
25 Weave Mamba Fusion: Global Cross-Scale Interaction for Lightweight Face Detection 提出Weave Mamba Fusion以解决轻量级人脸检测中的跨尺度依赖问题 Mamba state space model ✅
26 Talk Like You: Imitating How You Speak in Real-Time Talking Head Generation 提出TalkLikeYou以解决个性化口语习惯模仿问题 imitation learning flow matching ✅

🔬 支柱三：空间感知与语义 (Perception & Semantics) (6 篇)

#题目一句话要点标签🔗⭐
27 Wiring Matters: Injection Topology and Initialization of Affordance Heads in Vision-Language-Action Policies 提出有效的接入拓扑与初始化方法以提升视觉-语言-动作策略的性能 affordance vision-language-action VLA
28 GS-Pool: Object-Level Change Detection in 3D Gaussian Splatting 提出GS-Pool以解决3D高斯点云中的物体级变化检测问题 3D gaussian splatting 3DGS gaussian splatting
29 MaRO-GS: Mask-Robust Object-Centric Gaussian Splatting from Inconsistent Multi-view Masks 提出MaRO-GS以解决多视角不一致掩码下的3D重建问题 3DGS gaussian splatting splatting
30 Less Context, Better Geometry: Masked Geometric Encoder for Robust 3D Foundation Models 提出Masked Geometric Encoder以解决3D重建中的上下文不足问题 3D reconstruction foundation model
31 VGGT-Bridge: Beyond Sequential Pose Graphs via Coarse-Stride Skip Edges 提出VGGT-Bridge以解决长序列重建中的漂移问题 3D reconstruction VGGT
32 FrontVeg V2: A Training-Free Software Framework for Foreground-Aware Zero-Shot Plant Trait Segmentation in High-Resolution Images of Trellised Crops 提出FrontVeg V2以解决高分辨率作物图像中的植物特征分割问题 depth estimation monocular depth Depth Anything

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
33 KineWorld: Action-Induced Transport Fields for Embodied World Modeling 提出KineWorld以解决现有动作条件世界模型不足问题 manipulation bi-manual bimanual manipulation
34 TAPDreamer: Transferable Adversarial Patches for World Action Models 提出TAPDreamer以解决世界动作模型的视觉输入脆弱性问题 manipulation world model world models
35 SPIN: Image Immunization Against Diffusion Editing via Single-Step Projection in Stochastic Neighborhoods 提出SPIN框架以解决图像免疫对抗扩散编辑问题 manipulation

🔬 支柱八：物理动画 (Physics-based Animation) (2 篇)

#题目一句话要点标签🔗⭐
36 ChronoWorld: Camera-Controlled Consistent 4D World Generation via Spatiotemporal Cues and Geometric Reflections 提出ChronoWorld以解决4D场景生成一致性问题 spatiotemporal
37 A Spatiotemporal Semantic Importance-Guided Unified Compression and Editing Framework for AI-Generated Videos 提出统一压缩与编辑框架以解决AI生成视频存储与编辑问题 spatiotemporal

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

#题目一句话要点标签🔗⭐
38 Harnessing Multimodal Large Language Models for Training-Free Human-Object Interaction Detection 提出HarnessHOI以解决训练无关的人-物交互检测问题 human-object interaction HOI large language model

⬅️ 返回 cs.CV 首页 · 🏠 返回主页