Projects
Ongoing
This is a prestigious university research grant funded by YÖK that will work towards developing an AI-based coding and compression approach within the context of MPAI-CAE (Context-based Audio Enhancement) standard.
Past
This is a short and focused project that will investigate the use of RSMAs in measuring the directivity patterns of sound sources using a secondary source model.
The overarching aim of the project was to synthesise the sound field at a location in the recording volume using a limited number of RSMA recordings of the acoustic scene. Techniques were developed to that aim will include sound field extrapolation using spherical harmonic decompositions and object-based sparse approximations using a dictionary-based representation of the plane wave decomposition of the sound field. The planned research also includes work on transcoding the sound field for interactive binaural reproduction. Specifically, transcoding between interpolated higher-order Ambisonics and binaural audio will be investigated. The theoretical results of the project were be validated using real recordings.
This project aims to develop a spatial audio recording, coding, reproduction chain based on the analysis of sound scenes from recordings made with special microphone arrays. The project will investigate novel sound source localisation and separation algorithms and direct/diffuse separation from recordings; develop sound scene authoring tools and necessary methods for 3D and spatial sound source generation.
This project aims to design and develop open spherical microphone arrays to accurately measure acoustic intensity and use the acoustic intensity vectors for localising sound sources, even under highly reverberant environments.