System-to-user and user-to-system adaptation in binaural spatialisation Seminar

The binaural spatialisation technique allows to create 3-dimensional audio using a simple pair of headphones. It is widely known that everyone perceives spatial sound differently thanks to the particular shape of their ears, torso and head. High-quality simulations need therefore to be uniquely tailored to each individual in an effective and efficient manner. Various methods and techniques exist in order to improve the match between the HRTF (Head-Related Transfer Function) used for the rendering and the one specific to the listener. These can be mapped in two separate groups; the first one based on personalising the rendering system in order for it to match with the specific listener (system-to-user adaptation), and the second aiming at using auditory training to improve performances (e.g. localisation accuracy) of a given listener when using a non-individualised HRTF (user-to-system).

After a brief introduction to binaural spatialisation and to the most commonly known challenges with this technique, the presentation will focus on an overview of research studies aimed at tackling the matter of HRTF personalisation using the two approaches mentioned above. The talk then will close with some examples of practical applications of the techniques/methods presented, including a brief introduction to the BEARS and SONICOM research projects, as well as an overview of the tools developed for and used in these experiments (e.g. the 3D Tune-In Toolkit).

Speaker information

Dr Lorenzo Picinali, Imperial College London

"I am a Reader in Audio Experience Design at Imperial College London. In the past years I have worked in Italy, France and UK on projects related with binaural spatialisation and spatial hearing, looking both at the theoretical side of things and at applications in areas. "

Link to MS Teams

Click here to join the meeting

Or call in (audio only)
+44 20 3794 0272,,31788261# United Kingdom, London