Project A1 - Assessment and modelling of human speech intelligibility and listening effort
In many situations human speech communication is considerably affected due to hearing impairment and interfering sounds. This project will assess speech intelligibility and listening effort and will compare the perceptual data to model predictions for evaluating sound presentation and speech enhancement methods developed in the CRC.
Existing prediction models will be extended to supra-threshold processing for predicting listening effort and will be revised in order to work non-intrusively (enabling processing mixed and enhanced signals). Furthermore a real-time model will be developed allowing for direct feedback of temporally resolved model predictions to adaptive speech enhancement algorithms enabling in-the-loop applications.