Die Informatik des Fachbereiches 3 der Universität Bremen Hier geht es zur Homepage der Verwaltung des Fachbereiches 3 der Universität Bremen Hier geht es zur Homepage der Informatik des Fachbereiches 3 der Universität Bremen Hier geht es zur Homepage der Mathematik des Fachbereiches 3 der Universität Bremen Hier geht es zur Homepage des Fachbereiches 3 der Universität Bremen Hier geht es zur Homepage der Universität Bremen


INHALT & PFAD:
Startseite Detail


Silent Speech: Reconstructing Speech from Sensor Data by Machine Learning



Datum: 11.07.2018
Uhrzeit: 11.00 Uhr s.t.

Ort: Cartesium Rotunde


Vortragende(r): Jose A. Gonzalez, Ph.D. (Department of Languages and Computer Sciences, University of Malaga, Spain)

++ NEUES DATUM: 11.07. ++

Total removal of the larynx is often required to treat laryngeal cancer: every year some 17,500 people in Europe and North America lose the ability to speak in this way. Current methods for restoring speech include the electro-larynx, which produces an unnatural, electronic voice, oesophageal (belching) speech, which is difficult to learn, and fistula valve speech, which is considered to be the current gold standard but requires regular hospital visits for valve replacement and produces a masculine voice unpopular with female patients. All these methods sacrifice the patient's spoken identity.
In this talk I will present a technique which has the potential to restore the power of speech by sensing movement of the remaining speech articulators and using machine learning algorithms to derive a transformation which converts this sensor data into an acoustic signal - 'Silent Speech'. The sensing technique, developed by our collaborators at the University of Hull and called 'Permanent Magnetic Articulography', involves attaching small, unobtrusive magnets to the lips and tongue and monitoring changes in the magnetic field induced by their movement. I report experiments with several machine learning techniques and show that the Silent Speech generated, which may be delivered in real time, is intelligible and sounds natural. The identity of the speaker is recognisable.

Biografie

Jose A. Gonzalez is a lecturer in the Department of Languages and Computer Sciences, University of Malaga, Spain. He received the B.Sc. and Ph.D. degrees in Computer Science, both from the University of Granada, Spain, in 2006 and 2013, respectively. During his Ph.D. he did two research visits at the Speech and Hearing Research Group, University of Sheffield, U.K, to study missing data approaches for noise robust speech recognition. From 2013 to 2018 he was a Research Associate at the University of Sheffield working in clinical applications of speech technology.
He has co-/authored more than 60 international articles published in books, journals, and proceedings. He has received several scientific awards for his work, including AELFA-IF 2018 best paper award, BioDevices 2018 best paper award, BioSignals 2015 best paper award, Eusipco 2014 best student paper award, RTTH best journal paper award and the AVIOS Speech Application Contest 2007/2008 award.



Ansprechpartner(in) / Einladende(r):
Professorin Tanja Schultz


Download:




zurück  

Seitenanfang  -  Impressum Zuletzt geändert durch: jungmann [b]   05.07.2018 Admin-Login