About the SAP Laboratory

Speech and audio processing research in the Communications and Signal Processing Group at Imperial College London is addressing the fundamental science of speech and audio processing as well as technology applications particularly in telecoms and audio interfaces.

Recent topic areas include echo cancellation, dereverberation, speech enhancement, SIMO MIMO acoustic system identification, speech recognition, blind source separation extraction, speaker verification, speech analysis, Digital Audio Broadcast (DAB).

Current Projects

Hearing Aids

Developing speech enhancement algorithms that jointly enhance the speech received by the two ears


Robot Audition

We provide autonomous agents with a situational awareness of sound sources and reflectors in the acoustic environment.


IEEE-AASP Challenge on Localization & Tracking (LOCATA)

Open-source data corpus and benchmarking challenge for sound source localization & tracking algorithms.


Audio Diarization

We develop novel algorithms to address the question “who spoke when?”


Blind System Identification

We extract information from multichannel systems for dereverberation and source recovery.

Acoustic Imaging

We are investigating methods to characterise rough surfaces from acoustic reflections.



SAP papers at WASPAA 2019

We are delighted that the following papers have been accepted for WASPAA 2019 in New Paltz, New York, U.S.A in October 2019. V. Neo, C. Evers and P. Naylor: Speech Enhancement Using Polynomial Eigenvalue Decomposition. In: Proc. IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), 2019. A. Hogg, C. Evers and

SAP paper at EUSIPCO 2019

We are delighted that the following paper has been accepted for EUSIPCO 2019 in A Coruña, Spain in September 2019. D. Sharma, A. Hogg, Y. Wang, A. Nour-Eldin and P. Naylor: Non-Intrusive POLQA Estimation of Speech Quality using Recurrent Neural Networks. In: Proc. European Signal Process. Conf. (EUSIPCO), 2019. Looking forward to seeing you in A

SAP papers at ICASSP 2019

We are delighted that the following papers have been accepted for ICASSP 2019 in Brighton, UK in May 2019. A. Hogg, C. Evers and P. Naylor: Speaker Change Detection Using Fundamental Frequency With Application To Multi-talker Segmentation. In: Proc. IEEE Intl. Conf. on Acoustics, Speech and Signal Processing (ICASSP), 2019. V. Neo and P. Naylor: Second Order Sequential