Awesome Robotics Manipulation · full_paper

CLAMP: Contrastive Learning f or 3D Multi-View Action-Conditioned Robotic Manipulation Pretraining

作者：Krzysztof Choromanski, Sandy Huang, Connor Schenck · 单位：Google DeepMind University of Southern California · 会议/期刊：arXiv · 日期：2026-01-31 · 来源：Low-Level Learning-Based Action Modelling / Latent Learning / Pretrained Latent Learning

三维表征潜变量学习基础模型机器人学习操作
新标签打开 AlphaXiv新标签打开 Paper
CLAMP: Contrastive Learning f or 3D Multi-View Action-Conditioned Robotic Manipulation Pretraining figure
AlphaXiv 中文论文页面（可滚动查看）