Saved in:
Bibliographic Details
Main Authors: Qi, Han, Chen, Changhe, Yang, Heng
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.16053
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914376995307520
author Qi, Han
Chen, Changhe
Yang, Heng
author_facet Qi, Han
Chen, Changhe
Yang, Heng
contents A key requirement for generalist robots is compositional generalization - the ability to combine atomic skills to solve complex, long-horizon tasks. While prior work has primarily focused on synthesizing a planner that sequences pre-learned skills, robust execution of the individual skills themselves remains challenging, as visuomotor policies often fail under distribution shifts induced by scene composition. To address this, we introduce a scene graph-based representation that focuses on task-relevant objects and relations, thereby mitigating sensitivity to irrelevant variation. Building on this idea, we develop a scene-graph skill learning framework that integrates graph neural networks with diffusion-based imitation learning, and further combine "focused" scene-graph skills with a vision-language model (VLM) based task planner. Experiments in both simulation and real-world manipulation tasks demonstrate substantially higher success rates than state-of-the-art baselines, highlighting improved robustness and compositional generalization in long-horizon tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2509_16053
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Compose by Focus: Scene Graph-based Atomic Skills
Qi, Han
Chen, Changhe
Yang, Heng
Robotics
Artificial Intelligence
A key requirement for generalist robots is compositional generalization - the ability to combine atomic skills to solve complex, long-horizon tasks. While prior work has primarily focused on synthesizing a planner that sequences pre-learned skills, robust execution of the individual skills themselves remains challenging, as visuomotor policies often fail under distribution shifts induced by scene composition. To address this, we introduce a scene graph-based representation that focuses on task-relevant objects and relations, thereby mitigating sensitivity to irrelevant variation. Building on this idea, we develop a scene-graph skill learning framework that integrates graph neural networks with diffusion-based imitation learning, and further combine "focused" scene-graph skills with a vision-language model (VLM) based task planner. Experiments in both simulation and real-world manipulation tasks demonstrate substantially higher success rates than state-of-the-art baselines, highlighting improved robustness and compositional generalization in long-horizon tasks.
title Compose by Focus: Scene Graph-based Atomic Skills
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2509.16053