• Classified by Topic • Classified by Publication Type • Sorted by Date • Sorted by First Author Last Name • Classified by Funding Source •
GACL: Grounded Adaptive Curriculum Learning with Active Task and Performance Monitoring.
Linji Wang, Zifan
Xu, Peter Stone, and Xuesu Xiao.
In
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), October 2025.
Curriculum learning has emerged as a promising approach for training complexrobotics tasks, yet current applications predominantly rely on manually designedcurricula, which demand significant engineering effort and can suffer fromsubjective and suboptimal human design choices. While automated curriculumlearning has shown success in simple domains like grid worlds and games wheretask distributions can be easily specified, robotics tasks present uniquechallenges: they require handling complex task spaces while maintaining relevanceto target domain distributions that are only partially known through limitedsamples. To this end, we propose Grounded Adaptive Curriculum Learning (GACL), aframework specifically designed for robotics curriculum learning with three keyinnovations: (1) a task representation that consistently handles complex robottask design, (2) an active performance tracking mechanism that allows adaptivecurriculum generation appropriate for the robot's current capabilities, and (3) agrounding approach that maintains target domain relevance through alternatingsampling between reference and synthetic tasks. We validate GACL on wheelednavigation in constrained environments and quadruped locomotion in challenging 3Dconfined spaces, achieving 6.8 and 6.1 percent higher success rates, respectively, thanstate-of-the-art methods in each domain.
@InProceedings{linji_wang_iros2025,
author = {Linji Wang and Zifan Xu and Peter Stone and Xuesu Xiao},
title = {GACL: Grounded Adaptive Curriculum Learning with Active Task and Performance Monitoring},
booktitle = {IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year = {2025},
month = {October},
location = {Hangzhou, China},
abstract = {Curriculum learning has emerged as a promising approach for training complex
robotics tasks, yet current applications predominantly rely on manually designed
curricula, which demand significant engineering effort and can suffer from
subjective and suboptimal human design choices. While automated curriculum
learning has shown success in simple domains like grid worlds and games where
task distributions can be easily specified, robotics tasks present unique
challenges: they require handling complex task spaces while maintaining relevance
to target domain distributions that are only partially known through limited
samples. To this end, we propose Grounded Adaptive Curriculum Learning (GACL), a
framework specifically designed for robotics curriculum learning with three key
innovations: (1) a task representation that consistently handles complex robot
task design, (2) an active performance tracking mechanism that allows adaptive
curriculum generation appropriate for the robot's current capabilities, and (3) a
grounding approach that maintains target domain relevance through alternating
sampling between reference and synthetic tasks. We validate GACL on wheeled
navigation in constrained environments and quadruped locomotion in challenging 3D
confined spaces, achieving 6.8 and 6.1 percent higher success rates, respectively, than
state-of-the-art methods in each domain.
},
}
Generated by bib2html.pl (written by Patrick Riley ) on Wed Aug 06, 2025 21:39:47