Biometric signals compression with time- and subject-adaptive dictionary for wearable devices

This thesis work is dedicated to the design of a lightweight compression technique for the real-time processing of biomedical signals in wearable devices. The proposed approach exploits the unsupervised learning algorithm of the time-adaptive self-organizing map (TASOM) to create a subject-adaptive codebook applied to the vector quantization of a signal. The codebook is obtained and then dynamically refined in an online fashion, without requiring any prior information on the signal itself