ReVLA: Reverting Visual Domain Limitation of Robotic Foundation Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dey, Sombit, Zaech, Jan-Nico, Nikolov, Nikolay, Van Gool, Luc, Paudel, Danda Pani
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908371613908992
author Dey, Sombit
Zaech, Jan-Nico
Nikolov, Nikolay
Van Gool, Luc
Paudel, Danda Pani
author_facet Dey, Sombit
Zaech, Jan-Nico
Nikolov, Nikolay
Van Gool, Luc
Paudel, Danda Pani
contents Recent progress in large language models and access to large-scale robotic datasets has sparked a paradigm shift in robotics models transforming them into generalists able to adapt to various tasks, scenes, and robot modalities. A large step for the community are open Vision Language Action models which showcase strong performance in a wide variety of tasks. In this work, we study the visual generalization capabilities of three existing robotic foundation models, and propose a corresponding evaluation framework. Our study shows that the existing models do not exhibit robustness to visual out-of-domain scenarios. This is potentially caused by limited variations in the training data and/or catastrophic forgetting, leading to domain limitations in the vision foundation models. We further explore OpenVLA, which uses two pre-trained vision foundation models and is, therefore, expected to generalize to out-of-domain experiments. However, we showcase catastrophic forgetting by DINO-v2 in OpenVLA through its failure to fulfill the task of depth regression. To overcome the aforementioned issue of visual catastrophic forgetting, we propose a gradual backbone reversal approach founded on model merging. This enables OpenVLA -- which requires the adaptation of the visual backbones during initial training -- to regain its visual generalization ability. Regaining this capability enables our ReVLA model to improve over OpenVLA by a factor of 77\% and 66\% for grasping and lifting in visual OOD tasks. Comprehensive evaluations, episode rollouts and model weights are available on the ReVLA Page
format Preprint
id arxiv_https___arxiv_org_abs_2409_15250
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle ReVLA: Reverting Visual Domain Limitation of Robotic Foundation Models
Dey, Sombit
Zaech, Jan-Nico
Nikolov, Nikolay
Van Gool, Luc
Paudel, Danda Pani
Computer Vision and Pattern Recognition
Robotics
Recent progress in large language models and access to large-scale robotic datasets has sparked a paradigm shift in robotics models transforming them into generalists able to adapt to various tasks, scenes, and robot modalities. A large step for the community are open Vision Language Action models which showcase strong performance in a wide variety of tasks. In this work, we study the visual generalization capabilities of three existing robotic foundation models, and propose a corresponding evaluation framework. Our study shows that the existing models do not exhibit robustness to visual out-of-domain scenarios. This is potentially caused by limited variations in the training data and/or catastrophic forgetting, leading to domain limitations in the vision foundation models. We further explore OpenVLA, which uses two pre-trained vision foundation models and is, therefore, expected to generalize to out-of-domain experiments. However, we showcase catastrophic forgetting by DINO-v2 in OpenVLA through its failure to fulfill the task of depth regression. To overcome the aforementioned issue of visual catastrophic forgetting, we propose a gradual backbone reversal approach founded on model merging. This enables OpenVLA -- which requires the adaptation of the visual backbones during initial training -- to regain its visual generalization ability. Regaining this capability enables our ReVLA model to improve over OpenVLA by a factor of 77\% and 66\% for grasping and lifting in visual OOD tasks. Comprehensive evaluations, episode rollouts and model weights are available on the ReVLA Page
title ReVLA: Reverting Visual Domain Limitation of Robotic Foundation Models
topic Computer Vision and Pattern Recognition
Robotics
url https://arxiv.org/abs/2409.15250