Uncertainty analysis of deep neural network for classification of vulnerable road users using micro-doppler

Dubey A, Fuchs J, Reißland T, Weigel R, Lurz F (2020)


Publication Type: Conference contribution, Conference Contribution

Publication year: 2020

Publisher: Institute of Electrical and Electronics Engineers Inc.

Pages Range: 23-26

Conference Proceedings Title: 2020 IEEE Topical Conference on Wireless Sensors and Sensor Networks, WiSNeT 2020

Event location: San Antonio, TX US

ISBN: 9781728133553

DOI: 10.1109/WiSNeT46826.2020.9037574

Abstract

Unlike optical imaging, it's difficult to extract descriptive features from radar data for problems like classification of different targets. This paper takes the advantage of different neural network based architectures such as convolutional neural networks and long-short term memory to propose an end-to-end framework for classification of vulnerable road users. To make the network's prediction more reliable for automotive applications, a new concept of network uncertainty is introduced to the defined architectures. The signal processing tool chain described in this paper achieves higher accuracy than state-of-the-art algorithms while maintaining latency requirement for automotive applications.

Authors with CRIS profile

Related research project(s)

How to cite

APA:

Dubey, A., Fuchs, J., Reißland, T., Weigel, R., & Lurz, F. (2020). Uncertainty analysis of deep neural network for classification of vulnerable road users using micro-doppler. In 2020 IEEE Topical Conference on Wireless Sensors and Sensor Networks, WiSNeT 2020 (pp. 23-26). San Antonio, TX, US: Institute of Electrical and Electronics Engineers Inc..

MLA:

Dubey, Anand, et al. "Uncertainty analysis of deep neural network for classification of vulnerable road users using micro-doppler." Proceedings of the 2020 IEEE Topical Conference on Wireless Sensors and Sensor Networks, WiSNeT 2020, San Antonio, TX Institute of Electrical and Electronics Engineers Inc., 2020. 23-26.

BibTeX: Download