End-to-end grasping policies for human-in-the-loop robots via deep reinforcement learning

TL;DR

This paper introduces an end-to-end deep reinforcement learning approach for human-in-the-loop robot grasping, utilizing real human trajectories and a success model for transparency, addressing robustness issues in EMG-based systems.

Contribution

It presents a novel RL and IL based training method in a stochastic simulation environment with real trajectories, and a success model for understanding policy performance.

Findings

Effective grasping policy learned from real human trajectories.

Enhanced transparency through the success model.

Robustness improvements over traditional EMG-based methods.

Abstract

State-of-the-art human-in-the-loop robot grasping is hugely suffered by Electromyography (EMG) inference robustness issues. As a workaround, researchers have been looking into integrating EMG with other signals, often in an ad hoc manner. In this paper, we are presenting a method for end-to-end training of a policy for human-in-the-loop robot grasping on real reaching trajectories. For this purpose we use Reinforcement Learning (RL) and Imitation Learning (IL) in DEXTRON (DEXTerity enviRONment), a stochastic simulation environment with real human trajectories that are augmented and selected using a Monte Carlo (MC) simulation method. We also offer a success model which once trained on the expert policy data and the RL policy roll-out transitions, can provide transparency to how the deep policy works and when it is probably going to fail.