Machine-learning-based algorithms for radar applications
Loading...
Author
Chicco, F.
Piguet, Y.
Dia, M.
Dietler, S.
Manetakis, K.
Höchemer, M.
Maamari, N.
Kafi, G.
Dal Fabbro, P. A.
DOI
Abstract
This article presents how the use of machine-learning (ML) becomes a key enabler for radar sensor processing in two different workstreams: (1) hand gesture recognition and (2) cross-modal supervision and fusion with cameras for object detection. Radar’s robustness to lighting and environmental conditions, combined with its privacy benefits, drives the need for ML pipelines that separate clutter and imperfect motion while remaining efficient for embedded radar platforms. The results presented show that ML is a very promising technology to complement CSEM radar hardware platform for advanced radar sensing.
Publication Reference
CSEM Scientific and Technical Report 2025, p. 97–98
Year
2025