Saved in:
Bibliographic Details
Main Authors: Miao, Zhuochen, Lv, Jun, Fang, Hongjie, Jin, Yang, Lu, Cewu
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2506.21057
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916811498323968
author Miao, Zhuochen
Lv, Jun
Fang, Hongjie
Jin, Yang
Lu, Cewu
author_facet Miao, Zhuochen
Lv, Jun
Fang, Hongjie
Jin, Yang
Lu, Cewu
contents Imitation learning has emerged as a powerful paradigm in robot manipulation, yet its generalization capability remains constrained by object-specific dependencies in limited expert demonstrations. To address this challenge, we propose knowledge-driven imitation learning, a framework that leverages external structural semantic knowledge to abstract object representations within the same category. We introduce a novel semantic keypoint graph as a knowledge template and develop a coarse-to-fine template-matching algorithm that optimizes both structural consistency and semantic similarity. Evaluated on three real-world robotic manipulation tasks, our method achieves superior performance, surpassing image-based diffusion policies with only one-quarter of the expert demonstrations. Extensive experiments further demonstrate its robustness across novel objects, backgrounds, and lighting conditions. This work pioneers a knowledge-driven approach to data-efficient robotic learning in real-world settings. Code and more materials are available on https://knowledge-driven.github.io/.
format Preprint
id arxiv_https___arxiv_org_abs_2506_21057
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Knowledge-Driven Imitation Learning: Enabling Generalization Across Diverse Conditions
Miao, Zhuochen
Lv, Jun
Fang, Hongjie
Jin, Yang
Lu, Cewu
Robotics
Imitation learning has emerged as a powerful paradigm in robot manipulation, yet its generalization capability remains constrained by object-specific dependencies in limited expert demonstrations. To address this challenge, we propose knowledge-driven imitation learning, a framework that leverages external structural semantic knowledge to abstract object representations within the same category. We introduce a novel semantic keypoint graph as a knowledge template and develop a coarse-to-fine template-matching algorithm that optimizes both structural consistency and semantic similarity. Evaluated on three real-world robotic manipulation tasks, our method achieves superior performance, surpassing image-based diffusion policies with only one-quarter of the expert demonstrations. Extensive experiments further demonstrate its robustness across novel objects, backgrounds, and lighting conditions. This work pioneers a knowledge-driven approach to data-efficient robotic learning in real-world settings. Code and more materials are available on https://knowledge-driven.github.io/.
title Knowledge-Driven Imitation Learning: Enabling Generalization Across Diverse Conditions
topic Robotics
url https://arxiv.org/abs/2506.21057