Sarthak Mangla will present "IndivHear: an Individualized Adaptive Deep Learning-based Hearing Aid" for the Department of Speech, Language, and Hearing Sciences' Seminars in Hearing Research at Purdue (SHRP).
Meeting ID: 432 634 0458
Abstract: Prior to coming to Purdue, motivated by his grandmother’s experience with hearing loss and hearing aids, Sarthak developed an inexpensive smartphone-based individualized adaptive hearing aid using deep learning. For patients with mild hearing loss, hearing aids have limited user benefits due to the low accessibility of professional help, limitations in fitting and diagnostic procedures of hearing devices, and the tedious process of setting up hearing aids. For patients with moderate to profound hearing loss, hearing support is inadequate due to limitations in the individualized fitting and diagnostic procedures of hearing aids, inadequacies in the functional capabilities of existing hearing aids, and limited human-machine interfaces to steer the device based on the patient’s individual needs. In an attempt to solve these problems, Sarthak created an end-to-end solution that utilizes a novel data collection approach to train individualized deep learning networks as hearing aids. Additionally, he also developed fully automated and remote versions of traditional fitting and diagnostic procedures like Pure Tone Audiometry and Speech Audiometry.
