In a previous post of mine (which you should read now if you haven’t already) I explained how to create a MATLAB executable for the widely used PESQ algorithm. The main reason for wanting to do this was to save time when running a large amount of speech quality tests and the speed increases obtained from using a PESQ MEX function were amazing! At the time of writing that post, the MEX function was approximately 8 times faster than anything else online that I could find. In fact, it is still the fastest implementation I can find, however, I think there is room for improvement and I finally found some time to get it working. In this post, I will show step-by-step how you can compile the PESQ MEX function to accept audio vectors directly from MATLAB and, which, should give great speed increases when run on parallel cores.
Ever wondered if you could cancel someones voice without the need for a physical wall or partition? In this work presented at ICASSP 2017 in New Orleans, USA, we investigate the possibilities of cancelling speech over a loudspeaker wall. The method is not limited to speech, in-fact, it works much better for periodic signals as the non-stationarity of speech degrades the performance.
Blindly counting the number of speech sources (talkers) in a meeting room can be a difficult task. This paper was presented at HSCMA 2017 at the Google Offices in San Francisco and shows how using coherent-to-diffuse ratios could allow real-time source counting.
A hybrid loudspeaker system consisting of dynamic point sources and parametric loudspeaker models shows great results above, generally inevitable, soundfield aliasing frequencies.
This work was presented at APSIPA ASC 2016 in Jeju, Korea.
Presented at ICASSP 2016, we explore ways to improve speech privacy in multizone soundfield reproductions using intelligibility metrics, all while maintaining high speech quality.