[Audiatur et altera pars]
Enhancing speech recognition in challenging acoustic conditions is crucial for effective verbal communication, especially for individuals with hearing impairments.
Project IMBAS is a large-scale research project that investigates the intelligibility of different spoken languages. Communication channels often contain various maskers that degrade speech signals and reduce speech intelligibility. We hypothesize that each language is affected by these maskers in distinct ways.
To address this, Project IMBAS aims to improve multilingual speech intelligibility by creating a training dataset for use in Machine Learning (ML) and Deep Learning (DL) algorithms.