Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Managing Uncertainty of AI-based Perception for Autonomous Systems

 
: Henne, Maximilian; Schwaiger, Adrian; Weiß, Gereon

:
Fulltext urn:nbn:de:0011-n-5552619 (146 KByte PDF)
MD5 Fingerprint: bd1f47c02969da1b4e5a95e13839005b
Created on: 21.8.2019


Espinoza, H.:
Workshop on Artificial Intelligence Safety, AISafety 2019. Proceedings. Online resource : Co-located with the 28th International Joint Conference on Artificial Intelligence, IJCAI 2019; Macao, China, August 11-12, 2019
Macao, 2019 (CEUR Workshop Proceedings 2419)
http://ceur-ws.org/Vol-2419/
pp.57-60
Workshop on Artificial Intelligence Safety (AISafety) <2019, Macao>
International Joint Conference on Artificial Intelligence (IJCAI) <28, 2019, Macao>
English
Conference Paper, Electronic Publication
Fraunhofer ESK ()
uncertainty estimation neural networks; uncertainty estimation; neural network; perception; safety autonomous system; safety; autonomous system; Bayesian neural network; BNN; dynamic dependability management; artificial intelligence; AI

Abstract
With the advent of autonomous systems, machine perception is a decisive safety-critical part to make such systems become reality. However, presently used AI-based perception does not meet the required reliability for usage in real-world systems beyond prototypes, as for autonomous cars. In this work, we describe the challenge of reliable perception for autonomous systems. Furthermore, we identify methods and approaches to quantify the uncertainty of AI-based perception. Along with dynamic management of the safety, we show a path to how uncertainty information can be utilized for the perception, so that it will meet the high dependability demands of life-critical autonomous systems.

: http://publica.fraunhofer.de/documents/N-555261.html