Detection of speech events and speaker characteristics through photo-plethysmographic signal neural processing
This work addresses the problem of extracting biometric and speech information from noisy, low-cost PPG sensors for applications in wearable technology, though it appears incremental as it builds on prior proposals for using PPG beyond heart monitoring.
The authors tackled the problem of detecting speech events and speaker characteristics like gender and identity from photoplethysmogram (PPG) signals using end-to-end convolutional neural networks, achieving AUCs of 79% for gender verification, 89.0% for person verification, and around 69% for speech detection.
The use of photoplethysmogram signal (PPG) for heart and sleep monitoring is commonly found nowadays in smartphones and wrist wearables. Besides common usages, it has been proposed and reported that person information can be extracted from PPG for other uses, like biometry tasks. In this work, we explore several end-to-end convolutional neural network architectures for detection of human's characteristics such as gender or person identity. In addition, we evaluate whether speech/non-speech events may be inferred from PPG signal, where speech might translate in fluctuations into the pulse signal. The obtained results are promising and clearly show the potential of fully end-to-end topologies for automatic extraction of meaningful biomarkers, even from a noisy signal sampled by a low-cost PPG sensor. The AUCs for best architectures put forward PPG wave as biological discriminant, reaching $79\%$ and $89.0\%$, respectively for gender and person verification tasks. Furthermore, speech detection experiments reporting AUCs around $69\%$ encourage us for further exploration about the feasibility of PPG for speech processing tasks.