Human Preference-Based Learning for High-dimensional Optimization of Exoskeleton Walking Gaits

03/13/2020
by   Maegan Tucker, et al.
0

Understanding users' gait preferences of a lower-body exoskeleton requires optimizing over the high-dimensional gait parameter space. However, existing preference-based learning methods have only explored low-dimensional domains due to computational limitations. To learn user preferences in high dimensions, this work presents LineCoSpar, a human-in-the-loop preference-based framework that enables optimization over many parameters by iteratively exploring one-dimensional subspaces. Additionally, this work identifies gait attributes that characterize broader preferences across users. In simulations and human trials, we empirically verify that LineCoSpar is a sample-efficient approach for high-dimensional preference optimization. Our analysis of the experimental data reveals a correspondence between human preferences and objective measures of dynamic stability, while also highlighting inconsistencies in the utility functions underlying different users' gait preferences. This has implications for exoskeleton gait synthesis, an active field with applications to clinical use and patient rehabilitation.

READ FULL TEXT

page 1

page 5

page 6

research
09/26/2019

Preference-Based Learning for Exoskeleton Gait Optimization

This paper presents a personalized gait optimization framework for lower...
research
08/08/2022

POLAR: Preference Optimization and Learning Algorithms for Robotics

Parameter tuning for robotic systems is a time-consuming and challenging...
research
11/09/2020

ROIAL: Region of Interest Active Learning for Characterizing Exoskeleton Gait Preference Landscapes

Characterizing what types of exoskeleton gaits are comfortable for users...
research
11/10/2020

Preference-Based Learning for User-Guided HZD Gait Generation on Bipedal Walking Robots

This paper presents a framework that unifies control theory and machine ...
research
01/09/2018

Sequential Preference-Based Optimization

Many real-world engineering problems rely on human preferences to guide ...
research
10/10/2018

Batch Active Preference-Based Learning of Reward Functions

Data generation and labeling are usually an expensive part of learning f...
research
03/01/2019

Introducing Super Pseudo Panels: Application to Transport Preference Dynamics

We propose a new approach for constructing synthetic pseudo-panel data f...

Please sign up or login with your details

Forgot password? Click here to reset