Quantifying Assistive Robustness Via the Natural-Adversarial Frontier
- URL: http://arxiv.org/abs/2310.10610v1
- Date: Mon, 16 Oct 2023 17:34:54 GMT
- Title: Quantifying Assistive Robustness Via the Natural-Adversarial Frontier
- Authors: Jerry Zhi-Yang He, Zackory Erickson, Daniel S. Brown, Anca D. Dragan
- Abstract summary: RIGID is a method for training adversarial human policies that trade off between minimizing robot reward and acting human-like.
On an Assistive Gym task, we use RIGID to analyze the performance of standard collaborative Reinforcement Learning.
We also compare the frontier RIGID identifies with the failures identified in expert adversarial interaction, and with naturally-occurring failures during user interaction.
- Score: 40.125563987538044
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: Our ultimate goal is to build robust policies for robots that assist people.
What makes this hard is that people can behave unexpectedly at test time,
potentially interacting with the robot outside its training distribution and
leading to failures. Even just measuring robustness is a challenge. Adversarial
perturbations are the default, but they can paint the wrong picture: they can
correspond to human motions that are unlikely to occur during natural
interactions with people. A robot policy might fail under small adversarial
perturbations but work under large natural perturbations. We propose that
capturing robustness in these interactive settings requires constructing and
analyzing the entire natural-adversarial frontier: the Pareto-frontier of human
policies that are the best trade-offs between naturalness and low robot
performance. We introduce RIGID, a method for constructing this frontier by
training adversarial human policies that trade off between minimizing robot
reward and acting human-like (as measured by a discriminator). On an Assistive
Gym task, we use RIGID to analyze the performance of standard collaborative
Reinforcement Learning, as well as the performance of existing methods meant to
increase robustness. We also compare the frontier RIGID identifies with the
failures identified in expert adversarial interaction, and with
naturally-occurring failures during user interaction. Overall, we find evidence
that RIGID can provide a meaningful measure of robustness predictive of
deployment performance, and uncover failure cases in human-robot interaction
that are difficult to find manually. https://ood-human.github.io.
Related papers
- Common (good) practices measuring trust in HRI [55.2480439325792]
Trust in robots is widely believed to be imperative for the adoption of robots into people's daily lives.
Researchers have been exploring how people trust robot in different ways.
Most roboticists agree that insufficient levels of trust lead to a risk of disengagement.
arXiv Detail & Related papers (2023-11-20T20:52:10Z) - Habitat 3.0: A Co-Habitat for Humans, Avatars and Robots [119.55240471433302]
Habitat 3.0 is a simulation platform for studying collaborative human-robot tasks in home environments.
It addresses challenges in modeling complex deformable bodies and diversity in appearance and motion.
Human-in-the-loop infrastructure enables real human interaction with simulated robots via mouse/keyboard or a VR interface.
arXiv Detail & Related papers (2023-10-19T17:29:17Z) - ImitationNet: Unsupervised Human-to-Robot Motion Retargeting via Shared Latent Space [9.806227900768926]
This paper introduces a novel deep-learning approach for human-to-robot motion.
Our method does not require paired human-to-robot data, which facilitates its translation to new robots.
Our model outperforms existing works regarding human-to-robot similarity in terms of efficiency and precision.
arXiv Detail & Related papers (2023-09-11T08:55:04Z) - Learning Vision-based Pursuit-Evasion Robot Policies [54.52536214251999]
We develop a fully-observable robot policy that generates supervision for a partially-observable one.
We deploy our policy on a physical quadruped robot with an RGB-D camera on pursuit-evasion interactions in the wild.
arXiv Detail & Related papers (2023-08-30T17:59:05Z) - SACSoN: Scalable Autonomous Control for Social Navigation [62.59274275261392]
We develop methods for training policies for socially unobtrusive navigation.
By minimizing this counterfactual perturbation, we can induce robots to behave in ways that do not alter the natural behavior of humans in the shared space.
We collect a large dataset where an indoor mobile robot interacts with human bystanders.
arXiv Detail & Related papers (2023-06-02T19:07:52Z) - Improving safety in physical human-robot collaboration via deep metric
learning [36.28667896565093]
Direct physical interaction with robots is becoming increasingly important in flexible production scenarios.
In order to keep the risk potential low, relatively simple measures are prescribed for operation, such as stopping the robot if there is physical contact or if a safety distance is violated.
This work uses the Deep Metric Learning (DML) approach to distinguish between non-contact robot movement, intentional contact aimed at physical human-robot interaction, and collision situations.
arXiv Detail & Related papers (2023-02-23T11:26:51Z) - Learning Latent Representations to Co-Adapt to Humans [12.71953776723672]
Non-stationary humans are challenging for robot learners.
In this paper we introduce an algorithmic formalism that enables robots to co-adapt alongside dynamic humans.
arXiv Detail & Related papers (2022-12-19T16:19:24Z) - Learning Representations that Enable Generalization in Assistive Tasks [45.62648124988644]
We focus on enabling generalization in assistive tasks in which the robot is acting to assist a user.
We find that sim2real methods that encode environment (or population) parameters and work well in tasks that robots do in isolation, do not work well in assistance.
arXiv Detail & Related papers (2022-12-05T18:59:16Z) - Evaluation of Performance-Trust vs Moral-Trust Violation in 3D
Environment [1.4502611532302039]
We aim to design an experiment to investigate the consequences of performance-trust violation and moral-trust violation in a search and rescue scenario.
We want to see if two similar robot failures, one caused by a performance-trust violation and the other by a moral-trust violation have distinct effects on human trust.
arXiv Detail & Related papers (2022-06-30T17:27:09Z) - Show Me What You Can Do: Capability Calibration on Reachable Workspace
for Human-Robot Collaboration [83.4081612443128]
We show that a short calibration using REMP can effectively bridge the gap between what a non-expert user thinks a robot can reach and the ground-truth.
We show that this calibration procedure not only results in better user perception, but also promotes more efficient human-robot collaborations.
arXiv Detail & Related papers (2021-03-06T09:14:30Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.