Long-horizon imitation learning for robotic manipulation
Title: Long-horizon imitation learning for robotic manipulation
DNr: Berzelius-2024-180
Project Type: LiU Berzelius
Principal Investigator: Yufei Duan <yufeidu@kth.se>
Affiliation: Kungliga Tekniska högskolan
Duration: 2024-05-02 – 2024-12-01
Classification: 10201
Keywords:

Abstract

This project aims to develop advanced algorithms and methodologies for long horizon manipulation tasks in autonomous robotics. The project focuses on enabling robots to perform complex sequences of actions over extended periods, which is critical for applications in manufacturing, healthcare, and domestic environments. The ultimate goal is to enhance the efficiency, adaptability, and decision-making capabilities of robots in dynamic and unstructured settings.