The LTS2-Acoustics is working on binaural audio synthesis for applications in hearing aid technologies. Our activities deal with the improvement of the naturalness and realism of the hearing experiences achieved with binaural spatialization. In particular, it is desirable to enhance the perception of a sound located inside a room. One possible track to achieve this consists in simulating an artificial reverberation, which will provide the acoustic cues likely to enhance the experience of the listener regarding the feeling of immersion in the scene and general feeling of naturalness.
Reverberation algorithms can be heavy in terms of computational cost, which is not ideal considering the limitations of the processors currently available in hearing aids. Nevertheless, using Fourier transform techniques and manipulations, it is possible to optimize the computational cost to synthesize reverberation of various room sizes.
This project aims at exploring the literature to find out about the possible algorithms to achieve this goal and the implementation of one (or more) of these technics in the framework of a frame-based real-time model mimicking the behavior of a hearing aid.
Profile: Electrical engineering, computer science
Prerequisites: Electroacoustique (BA5) or Signal processing (BA5) or Audio Engineering (MA1), experience with Matlab/Simulink
Content: Audio signal processing, Psychoacoustics, Matlab and Simulink implementation