Research Projects

The World Health Organization (WHO) estimates that 2.5 billion people will have some form of hearing loss by the year 2050. Further, the WHO estimates a significant rise in hearing impairment among young adults, which is caused explicitly by unsafe listening practices, including prolonged digital audio exposure. In addition to communication and learning difficulties, unattended hearing loss can affect cognition and mental health. Creating awareness among the young population on this upcoming health emergency, which the WHO terms as “new diabetics”, is of utmost importance. Most hearing losses are irreversible, and one of the most commonly employed rehabilitation methods for people with hearing loss is an assistive listening device, commonly referred to as a hearing aid. In the simplest form, a hearing aid has one or more microphones, a digital signal processing unit which processes the sound picked up by the microphones, a loudspeaker which plays the processed sound, which is usually also amplified to the ear of the wearer and a suitable power supply. Recent research has shown that one of the important consequences of impaired auditory capacity is increased listening effort. Listening effort refers to the increased cognitive resources required to process auditory information, particularly in noisy or unpredictable environments. Conventional hearing aids often amplify or distort sounds without accounting for the cognitive load placed on the user. This is particularly significant during complex auditory scenes involving multiple speakers or dynamic sound sources. This increase in listening effort has been shown to cause mental fatigue, reduced communication effectiveness, and social withdrawal. Therefore, understanding and quantifying listening effort from a cognitive and psychoacoustic perspective is essential for designing intelligent hearing aids that also support efficient, low-effort auditory processing. The psychoacoustic study will be followed by the development of algorithms for hearing aids, which use the insights from the psychoacoustic study to improve algorithms for hearing aids. Understanding speech in the presence of noise has been reported to be challenging for hearing aid users. Multi-microphone hearing aids, which use array signal processing principles to achieve spatial filtering using a suitable beamforming technique, have been recently developed. Most of these beamforming methods do not consider the head movement of the hearing aid user, which can affect the beamforming and spatial filtering performance. In this project, we aim to develop a multi-microphone hearing aid with beamforming algorithms that are immune to head movement. Using non-audio signals from other sensors in the hearing aid may also be used to pick up sounds from a direction the wearer wishes to focus on. The project also aims to develop new and improved methods for enhancing speech intelligibility for hearing aid users. Conventional and machine learning-based solutions, suitable for a low-power device like a hearing aid, will be explored to develop an appropriate speech intelligibility enhancement scheme. This may also need the development of a situation awareness algorithm, to identify the acoustic scene in which the hearing aid user is present and deploy a suitable noise reduction algorithm appropriate for the acoustic environment to provide the best possible noise reduction and speech intelligibility.