AVAESA: Adaptive VAE with Self-Attention and Learnable Signal Processing for Robust Radar-Based Heart Rate Estimation

• Novel AVAESA architecture with dual-stream I/Q processing and self-attention for enhanced temporal modeling • Adaptive signal processing reduces dependency on fixed filter parameters through learnable, signal-driven preprocessing • Improved cross-domain generalization maintains performance across diverse measurement conditions with statistical validation • Comprehensive evaluation against multiple baselines (CNN, LSTM, Bi-LSTM, TCN, VAE) establishes architectural design principles for contactless cardiac monitoring Non-contact heart rate monitoring using radar sensors offers significant advantages for healthcare and automotive applications by preserving privacy while enabling continuous physiological assessment. Current Variational Autoencoder (VAE) approaches for radar-based vital sign monitoring, while superior to traditional neural networks, suffer from fixed preprocessing assumptions and inadequate temporal modeling that limit their generalization across diverse measurement conditions. This study introduces AVAESA (Adaptive VAE with Self-Attention and Learnable Signal Processing), a novel architecture that addresses these limitations through three key innovations: dual-stream in-phase/quadrature signal processing that preserves critical phase relationships, multi-head self-attention mechanisms for enhanced temporal dependency modeling, and adaptive signal preprocessing with learnable parameters that derive frequency bands and processing weights directly from input signal characteristics. The framework was evaluated on 1,920 measurements from 10 participants across 48 measurement scenarios (4 distances × 3 angles × 4 orientations), assessing cross-scenario robustness under measurement domain shift, with Polar H10 chest strap ground truth validation. Comprehensive comparison against multiple architectures (CNN, LSTM, Bi-LSTM, TCN, VAE) with statistical significance testing demonstrates substantial performance improvements, with mean absolute error reductions ranging from 17.3% under optimal conditions to 62.6% under challenging cross-domain generalization scenarios. AVAESA maintains high accuracy (correlation coefficient > 0.86, R 2 > 0.84) even under extreme domain shift conditions where baseline approaches exhibit degraded performance, demonstrating potential for contactless cardiac monitoring systems across diverse measurement environments through improved cross-scenario robustness.