• • GenENet reduces electrode count from 32 to 6 channels while maintaining equivalent performance, cutting hardware complexity by 81.25% and associated power consumption for data acquisition and transmission.
• • The sensor uses PDMS substrate, EGaln interconnects, and PEDOT:PSS hydrogel, achieving significantly lower skin-contact impedance than standard dry electrodes, ensuring high SNR even under mechanical strain.
• • The masked autoencoder architecture learns anatomical synergies from high-density data, enabling accurate reconstruction of full 32-channel patterns from sparse 6-channel inputs, thus enabling real-time processing with reduced data throughput.
• • The approach demonstrates a paradigm shift from hardware-intensive to software-defined wearable systems, potentially enabling longer battery life and improved user comfort for continuous health monitoring and HMI applications.