By using the Aculab site, you agree with our use of cookies.

Detecting synthetic speech and identifying speakers: presentations on forensic speech science from the University of York

Our PhD student at the University of York, Elliot Holmes, has been keeping busy lately!

Already this year, he's presented his work on speech forensics at both the Associazione Italiana di Scienze della Voce conference, AISV 2021, and the Manchester Forum in Linguistics (mFiL 2021).

You can download his posters, “Using Phonetic Theory to Improve Automatic Speaker Recognition” and “Using Phonetic Theory to Recognise Synthetic Speech”, here:

AISV 2021 poster
mFiL 2021 poster

In them, Elliot discusses his methods for choosing the most useful phonetic features for two related task domains: speaker identification and detection of synthetic speech. These topics are increasingly important for real-world speech forensics and voice biometrics.

To find out more, check out the conference websites:

AISV Conference 2021:
mFiL 2021: