Composition of Robot Motions based on the Concept of Deep Predictive Learning

Abstract

A deep learning-based approach can generalize model performance while reducing feature design costs by learning end-to-end environment recognition and motion generation. However, the process incurs huge training data collection costs and time and human resources for trial-and-error when involving physical contact with robots. Therefore, we propose "deep predictive learning," a motion learning concept that assumes imperfections in the predictive model and minimizes the prediction error with the real-world situation. Deep predictive learning is inspired by the "free energy principle and predictive coding theory," which explains how living organisms behave to minimize the prediction error between the real world and the brain. Robots predict near-future situations based on sensorimotor information and generate motions that minimize the gap with reality. The robot can flexibly perform tasks in unlearned situations by adjusting its motion in realtime while considering the gap between learning and reality. This paper describes the concept of deep predictive learning, its implementation, and examples of its application to real robots. The code and document are available at https: //ogata-lab.github.io/eipl-docs

Enhancement of Long-Horizon Task Planning via Active and Passive Modification in Large Language Model

“…Thus, active modification is expected to contribute to reducing the design cost of natural language instructions. This is also important from an active inference in robotics fields 30 . This study aims to further refine action plans obtained from active modification through dialogue by passive modification, and few studies have addressed these issues for long-horizon robot tasks.…”

Cooking Robot Mastery: Learning Online Motion Generation with Active Perception

“…We introduce an attention mechanism that can weigh the modality in a predictive deep learning model considering the reliability and importance at that time. While previous research has used attention mechanisms to detect gaze points in images or for visual explanation [26,27,28], we use the mechanism for multimodal data to map the importance of the information.…”