Data-Efficient Artificial Neural Networks with Gaussian Process Regression for 3D Visible Light Positioning
Paper i proceeding, 2021
A data-efficient artificial neural network based 3D VLP system using Gaussian process regression is proposed and experimentally demonstrated. Training data from the real environment can be reduced by >50% with negligible loss in positioning accuracy.
Författare
Weikang Zeng
South China Normal University
Huayang Chen
South China Normal University
Jiajia Chen
Chalmers, Elektroteknik, Kommunikation, Antenner och Optiska Nätverk
South China Normal University
Xuezhi Hong
South China Normal University
2021 Optical Fiber Communications Conference and Exhibition, OFC 2021 - Proceedings
9489805
9781943580866 (ISBN)
2021 Optical Fiber Communications Conference and Exhibition, OFC 2021
San Francisco, USA,
San Francisco, USA,
Ämneskategorier
Bioinformatik (beräkningsbiologi)
Signalbehandling
Datorseende och robotik (autonoma system)