Spatially Visual Perception for End-to-End Robotic Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Davies, Travis, Yan, Jiahuan, Chen, Xiang, Tian, Yu, Zhuang, Yueting, Huang, Yiqi, Hu, Luhui
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912133967511552
author Davies, Travis
Yan, Jiahuan
Chen, Xiang
Tian, Yu
Zhuang, Yueting
Huang, Yiqi
Hu, Luhui
author_facet Davies, Travis
Yan, Jiahuan
Chen, Xiang
Tian, Yu
Zhuang, Yueting
Huang, Yiqi
Hu, Luhui
contents Recent advances in imitation learning have shown significant promise for robotic control and embodied intelligence. However, achieving robust generalization across diverse mounted camera observations remains a critical challenge. In this paper, we introduce a video-based spatial perception framework that leverages 3D spatial representations to address environmental variability, with a focus on handling lighting changes. Our approach integrates a novel image augmentation technique, AugBlender, with a state-of-the-art monocular depth estimation model trained on internet-scale data. Together, these components form a cohesive system designed to enhance robustness and adaptability in dynamic scenarios. Our results demonstrate that our approach significantly boosts the success rate across diverse camera exposures, where previous models experience performance collapse. Our findings highlight the potential of video-based spatial perception models in advancing robustness for end-to-end robotic learning, paving the way for scalable, low-cost solutions in embodied intelligence.
format Preprint
id arxiv_https___arxiv_org_abs_2411_17458
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Spatially Visual Perception for End-to-End Robotic Learning
Davies, Travis
Yan, Jiahuan
Chen, Xiang
Tian, Yu
Zhuang, Yueting
Huang, Yiqi
Hu, Luhui
Computer Vision and Pattern Recognition
Artificial Intelligence
Robotics
Recent advances in imitation learning have shown significant promise for robotic control and embodied intelligence. However, achieving robust generalization across diverse mounted camera observations remains a critical challenge. In this paper, we introduce a video-based spatial perception framework that leverages 3D spatial representations to address environmental variability, with a focus on handling lighting changes. Our approach integrates a novel image augmentation technique, AugBlender, with a state-of-the-art monocular depth estimation model trained on internet-scale data. Together, these components form a cohesive system designed to enhance robustness and adaptability in dynamic scenarios. Our results demonstrate that our approach significantly boosts the success rate across diverse camera exposures, where previous models experience performance collapse. Our findings highlight the potential of video-based spatial perception models in advancing robustness for end-to-end robotic learning, paving the way for scalable, low-cost solutions in embodied intelligence.
title Spatially Visual Perception for End-to-End Robotic Learning
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Robotics
url https://arxiv.org/abs/2411.17458