arXiv Robotics — research abstracts· Jiahui Fu, Junyu Nan, Lingfeng Sun, Hongyu Li, Jianing Qian, Benjamin Yang, Yilun Du, Jennifer L. Barry, Kris Kitani, George Konidaris·· 3 days agoEditorial score65
NovaPlan: Zero-Shot Long-Horizon Manipulation via Closed-Loop Video Language Planning
NovaPlan: Zero-Shot Long-Horizon Manipulation via Closed-Loop Video Language Planning
Summary
NovaPlan is a hierarchical framework that enables robust, zero-shot long-horlicity manipulation by combining Vision-Language Models (VLMs) with video generation models. It systematically proposes, verifies, and repairs visual plans, using a hybrid geometric representation and continuous execution monitoring to achieve complex assembly and dexterous error recovery without task-specific training or demonstrations.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
Source:arXiv Robotics — research abstracts · arxiv.org