arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2109.09690cs.ROcs.AIcs.CVcs.LG
收藏

相信你的机器人!稀疏高斯过程神经网络的预测不确定性估计

Trust Your Robots! Predictive Uncertainty Estimation of Neural Networks with Sparse Gaussian Processes

German Aerospace Center (DLR)(德国航空航天中心);Karlsruhe Institute of Technology (KIT)(卡尔斯鲁厄理工学院);Technical University of Munich (TUM)(慕尼黑工业大学);ETH Zürich (ETHZ)(苏黎世联邦理工学院)

Jongseok Lee, Jianxiang Feng, Matthias Humt, Marcus G. Müller, Rudolph Triebel

摘要:

本文提出了一个概率框架,以获得可靠和快速的不确定性估计与深度神经网络(DNN)的预测。我们的主要贡献是DNN与稀疏高斯过程(GP)的实用和原则性组合。我们从理论上证明,DNN可以被看作是一个特殊的情况下,稀疏GP,即GP专家的混合物(MoE-GP),我们设计了一个学习算法,将派生的理论付诸实践。在实验中,从两个不同的机器人任务-逆动力学的机械手和目标检测的微型飞行器(MAV)-我们展示了我们的方法的有效性方面的预测不确定性,提高可扩展性,和运行时的效率Jetson TX 2。因此,我们认为,我们的方法可以铺平道路,可靠和快速的机器人学习系统的不确定性意识。

英文摘要:

This paper presents a probabilistic framework to obtain both reliable and fast uncertainty estimates for predictions with Deep Neural Networks (DNNs). Our main contribution is a practical and principled combination of DNNs with sparse Gaussian Processes (GPs). We prove theoretically that DNNs can be seen as a special case of sparse GPs, namely mixtures of GP experts (MoE-GP), and we devise a learning algorithm that brings the derived theory into practice. In experiments from two different robotic tasks -- inverse dynamics of a manipulator and object detection on a micro-aerial vehicle (MAV) -- we show the effectiveness of our approach in terms of predictive uncertainty, improved scalability, and run-time efficiency on a Jetson TX2. We thus argue that our approach can pave the way towards reliable and fast robot learning systems with uncertainty awareness.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑