CLAM: Continuous Latent Action Models for Robot Learning from Unlabeled Demonstrations
CLAM:基于未标记演示的连续潜在动作模型用于机器人学习
机构 * Department of Computer Science, University of Southern California(南加州大学计算机科学系) ; Department of Electrical and Computer Engineering, University of Southern California(南加州大学电气与计算机工程系)
专题命中 机器人学习 :robot learning(title);manipulation(abstract);分类 cs.RO、cs.AI、cs.LG
AI总结 CLAM通过连续潜在动作标签和联合训练动作解码器,有效解决复杂连续控制任务中未标记数据的学习问题,实现在DMControl和MetaWorld等基准及真实机器人上的性能提升。
Comments Latent Action Models, Self-supervised Pretraining, Learning from Videos
Journal ref IEEE/RSJ International Conference on Intelligent Robots and Systems 2026