A Bio-Inspired Mechanism for Learning Robot Motion From Mirrored Human Demonstrations

Abstract

Different learning modes and mechanisms allow faster and better acquisition of skills as widely studied in humans and many animals. Specific neurons, called mirror neurons, are activated in the same way whether an action is performed or simply observed. This suggests that observing others performing movements allows to reinforce our motor abilities. This implies the presence of a biological mechanism that allows creating models of others' movements and linking them to the self-model for achieving mirroring. Inspired by such ability, we propose to build a map of movements executed by a teaching agent and mirror the agent's state to the robot's configuration space. Hence, in this study, a neural network is proposed to integrate a motor cortex-like differential map transforming motor plans from task-space to joint-space motor commands and a static map correlating joint-spaces of the robot and a teaching agent. The differential map is developed based on spiking neural networks while the static map is built as a self-organizing map. The developed neural network allows the robot to mirror the actions performed by a human teaching agent to its own joint-space and the reaching skill is refined by the complementary examples provided. Hence, experiments are conducted to quantify the improvement achieved thanks to the proposed learning approach and control scheme.

Analysis of the value of artificial intelligence combined with musculoskeletal ultrasound in the differential diagnosis of pain rehabilitation of scapulohumeral periarthritis

“…[11] Currently, the robot doctor with deep learning function has not only started practicing but also made promising progress. [12] To utilize artificial intelligence for disease diagnosis, artificial intelligence first needs to be able to perceive the surrounding environment like humans, particularly the status of the body, such as what a normal versus abnormal body entails, the presence of cancer in the body, or even minor changes in the body (i.e., a few cancerous cells in a very small biological environment). [11][12][13] This study enrolled 165 patients with periarthritis of the shoulder, who were admitted to our hospital from January 2020 to January 2022, to explore the value of artificial intelligence combined with musculoskeletal ultrasound in the differential diagnosis of pain rehabilitation of periarthritis of the shoulder.…”

“…[ 11 ] Currently, the robot doctor with deep learning function has not only started practicing but also made promising progress. [ 12 ] To utilize artificial intelligence for disease diagnosis, artificial intelligence first needs to be able to perceive the surrounding environment like humans, particularly the status of the body, such as what a normal versus abnormal body entails, the presence of cancer in the body, or even minor changes in the body (i.e., a few cancerous cells in a very small biological environment). [ 11 – 13 ]…”

A structured prediction approach for robot imitation learning

“…To achieve this goal, the robot should behave as the demonstrator in response to a query input. It is then critical to find a suitable mapping between the input and output values based on the collected demonstrations (Zahra et al, 2022). This classical view on imitation learning is reminiscent of the objective of supervised learning: To find an input/output function given input/output pairs.…”

Ultrasound-Guided Assistive Robots for Scoliosis Assessment with Optimization-based Control and Variable Impedance

“…In this section, we study the problem of regulating the impedance gains, namely the stiffness and damping matrices K D u , K P u , K D R , and K P R of (7), to guarantee safe and reliable interaction with the human patients. To this end, we consider to retrieve the desired impedance gains by resorting to the general framework of learning by demonstration or imitation learning since it provides a straightforward and intuitive manner to transfer human motion skills to robots [27]. Specifically, we propose to leverage the techniques from probabilistic imitation learning framework, where we can obtain variability from multiple human demonstrations to regulate the robot Fig.…”