Privacy-Preserving Localization and Social Distance Monitoring with Low-Resolution Thermal Imaging and Deep Learning
Publikation: Beiträge in Zeitschriften › Zeitschriftenaufsätze › Forschung › begutachtet
Standard
in: Procedia CIRP, Jahrgang 130, 01.01.2024, S. 355-361.
Publikation: Beiträge in Zeitschriften › Zeitschriftenaufsätze › Forschung › begutachtet
Harvard
APA
Vancouver
Bibtex
}
RIS
TY - JOUR
T1 - Privacy-Preserving Localization and Social Distance Monitoring with Low-Resolution Thermal Imaging and Deep Learning
AU - Perov, Andrei
AU - Heger, Jens
N1 - Conference code: 57
PY - 2024/1/1
Y1 - 2024/1/1
N2 - This study introduces a novel approach to leverage low-power, low-resolution infrared sensors for detailed people tracking in manufacturing settings. We curated a dataset including a diverse range of interactions labeled for multiple-person localization and social distance violation tasks. Our methodology uses a combination of convolutional and recurrent neural networks to interpret spatiotemporal data. We demonstrate the capability of the novel image segmentation approach for human localization where we achieve 97.5 percent image-level accuracy. Also, we highlight the importance of interpolation and convolutional kernel selection for social distance tasks where we achieve 91 percent macro-averaged accuracy in 4 class scenarios.
AB - This study introduces a novel approach to leverage low-power, low-resolution infrared sensors for detailed people tracking in manufacturing settings. We curated a dataset including a diverse range of interactions labeled for multiple-person localization and social distance violation tasks. Our methodology uses a combination of convolutional and recurrent neural networks to interpret spatiotemporal data. We demonstrate the capability of the novel image segmentation approach for human localization where we achieve 97.5 percent image-level accuracy. Also, we highlight the importance of interpolation and convolutional kernel selection for social distance tasks where we achieve 91 percent macro-averaged accuracy in 4 class scenarios.
KW - Deep Learning
KW - Infrared Sensors
KW - Convolutional Neural Networks
KW - Facility Layout Planning
KW - Multiple Object Localization
KW - Engineering
UR - http://www.scopus.com/inward/record.url?scp=85213010738&partnerID=8YFLogxK
U2 - 10.1016/j.procir.2024.10.100
DO - 10.1016/j.procir.2024.10.100
M3 - Journal articles
VL - 130
SP - 355
EP - 361
JO - Procedia CIRP
JF - Procedia CIRP
SN - 2212-8271
T2 - 57th CIRP Conference on Manufacturing Systems - CIRP CMS 2024
Y2 - 29 May 2024 through 31 May 2024
ER -