Lee, JeongminKwon, TaesooShin, HyunjuLee, YoonsangHu, RuizhenCharalambous, Panayiotis2024-04-302024-04-302024978-3-03868-237-01017-4656https://doi.org/10.2312/egs.20241020https://diglib.eg.org/handle/10.2312/egs20241020We present an approach using deep reinforcement learning (DRL) to directly generate motion matching queries for longterm tasks, particularly targeting the reaching of specific locations. By integrating motion matching and DRL, our method demonstrates the rapid learning of policies for target location tasks within minutes on a standard desktop, employing a simple reward design. Additionally, we propose a unique hit reward and obstacle curriculum scheme to enhance policy learning in environments with moving obstacles.Attribution 4.0 International LicenseCCS Concepts: Computing methodologies → Motion processing; Motion path planningComputing methodologies → Motion processingMotion path planning10.2312/egs.202410204 pages