A 65 nm Multi-Modal Bayesian Inference Engine with 16.3 fJ/Sample Calibration-Free GRNG for Risk-Aware At-Home Skin Lesion Screening
This work addresses the need for privacy-preserving, energy-efficient, and risk-aware edge-AI for medical screening in uncontrolled at-home settings.
The paper presents a 65-nm risk-aware multimodal Bayesian inference engine for on-device skin lesion screening, achieving 1.4x improvement in equal-risk operating coverage, >1.5x robustness to perturbations, 5.5x process-variation resilience, and 1.8% balanced accuracy gain over unimodal Bayesian neural networks, with 16.3 fJ/sample efficiency.
We present a 65-nm risk-aware multimodal Bayesian inference engine for privacy-preserving, fully on-device skin lesion screening under uncontrolled at-home conditions. The proposed compute-in-memory architecture performs in-word Mixture-of-Gaussian sampling, improving uncertainty modeling beyond conventional unimodal Bayesian neural networks. This added probabilistic expressiveness increases equal-risk operating coverage by 1.4x, improves robustness to user-data perturbations by >1.5x, enhances process-variation resilience by 5.5x, and improves balanced accuracy by 1.8% over state-of-the-art unimodal Bayesian neural networks. Hardware robustness is further supported by calibration-free Gaussian random-number generation using complementary process variation, achieving 16.3 fJ/sample and 168.6 GSa/s/mm^2 efficiency. These results demonstrate a practical, energy-efficient, and risk-aware edge-AI solution for privacy-conscious medical screening.