Neuromorphic respiration rate detection using convolutional neural networks

Abstract

Ambient assisted living (AAL) technologies can significantly enhance autonomy for elderly and disabled individuals.

A key challenge is balancing effective monitoring with privacy, as conventional cameras may be perceived as intrusive. Neu-

romorphic cameras offer a privacy-preserving alternative by sensing only changes in brightness rather than full image frames

and operating at lower spatial resolution. This work investigates a non-invasive approach for estimating patients’ respiration

rate using neuromorphic cameras combined with convolutional neural networks.