Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Person Identification and Body Mass Index: A Deep Learning-Based Study on Micro-Dopplers

 
: Abdulatif, Sherif; Aziz, Fady; Armanious, Karim; Kleiner, Bernhard; Yang, Bin; Schneider, Urs

:

Institute of Electrical and Electronics Engineers -IEEE-:
IEEE Radar Conference, RadarConf 2019 : 22-26 April 2019, Westin Waterfront Hotel, Boston, Massachusetts, USA
Piscataway, NJ: IEEE, 2019
ISBN: 978-1-7281-1679-2
ISBN: 978-1-7281-1680-8
S.2317-2322
Radar Conference (RadarConf) <2019, Boston/Mass.>
Englisch
Konferenzbeitrag
Fraunhofer IPA ()
Body-Mass-Index; deep learning; Doppler radar; locomotion

Abstract
Obtaining a smart surveillance requires a sensing system that can capture accurate and detailed information for the human walking style. The radar micro-Doppler (μ- D) analysis is proved to be a reliable metric for studying human locomotions. Thus, μ- D signatures can be used to identify humans based on their walking styles. Additionally, the signatures contain information about the radar cross section (RCS) of the moving subject. This paper investigates the effect of human body characteristics on human identification based on their μ-D signatures. In our proposed experimental setup, a treadmill is used to collect μ- D signatures of 22 subjects with different genders and body characteristics. Convolutional autoencoders (CAE) are then used to extract the latent space representation from the μ- D signatures. It is then interpreted in two dimensions using t-distributed stochastic neighbor embedding (t-SNE). Our study shows that the body mass index (BMI) has a correlation with the μ-D signature of the walking subject. A 50-layer deep residual network is then trained to identify the walking subject based on the μ-D signature. We achieve an accuracy of 98% on the test set with high signal-to-noise-ratio (SNR) and 84% in case of different SNR levels.

: http://publica.fraunhofer.de/dokumente/N-585612.html