Skip to main navigation menu Skip to main content Skip to site footer

Vol. 11 No. 1 (2025)
Special Issue: Proceedings of CVIS 2025

Articles

A Physics-Informed Digital Twin Framework for Calibrated Sim-to-Real FMCW Radar Occupancy Estimation

https://doi.org/10.15353/jcvis.v11i1.10016

Published 2026-04-01

  • Huy Trinh
    • ,
  • Sebastian Ratto V.
    • ,
  • Elliot Creager
    • ,
  • George Shaker

Abstract

Learning robust radar perception models directly from real measurements is costly due to the need for controlled experiments, repeated calibration, and extensive annotation. This paper proposes a lightweight simulation-to-real (sim2real) framework that can enable reliable Frequency Modulated Continuous Wave (FMCW) radar occupancy detection and people counting using only a physics-informed geometric simulator plus a small unlabeled real calibration set. A calibrated domain randomization (CDR) step is introduced that aligns the global noise-floor statistics of simulated range--Doppler (RD) maps to those observed in real environments, while preserving discriminative micro-Doppler structure. Across real-world evaluations, ResNet18 models trained purely on CDR-adjusted simulation achieve 97% accuracy for occupancy detection and 72% accuracy for people counting, outperforming ray tracing baseline simulation and conventional random domain-randomization baselines.
PDF