Amelia Gully
Lecturer in Speech Science

Research

Overview

My research sits at the junction of phonetics and numerical acoustic modelling. I use MRI and other imaging technologies to capture the shape of the vocal tract, informing a 3D model of acoustic wave propagation within the airway. I am interested in how differences in vocal anatomy contribute to making a speaker’s voice unique, both for forensic speaker comparison applications and in order to inform personalised speech synthesis for people who have lost their voice.

Interests

• Speech synthesis
• Numerical acoustic modelling
• Speaker identification
• Morphoacoustics
• Augmentative and alternative communication

Projects

Anatomy, acoustics, and the individual: investigating inter-speaker vocal tract variation for forensic speaker comparison (2019-2023)

British Academy Postdoctoral Fellowship

Project page

This project makes use of cutting edge acoustic and morphometric techniques to understand how the unique physiology of a person's vocal tract affects the acoustics of their voice, with applications to forensic speaker identification. 

From swallowing to speech to singing: investigating the vocal tract using electromagnetic articulography and ultrasound (2019)

University of York Research Priming Fund

I am Co-I on this project which seeks to bring together voice researchers across the University of York to develop new collaborations making use of articulograph and ultrasound technology.

Voice and Identity: Source, Filter, Biometric (2015-2019)

Arts and Humanities Research Council

I joined this project as a Research Associate in 2018, working on developing and implementing an MRI and vocal tract modelling protocol to investigate the production of vocal profile analysis (VPA) settings for forensic phonetics applications.

Stop consonant synthesis using the digital waveguide mesh vocal tract model (2017)

World Universities Network Research Mobility Fund

I was awarded funding to visit Alberta Phonetics Laboratory and investigate using the three-dimensional dynamic digital waveguide mesh vocal tract model to simulate voiced stop consonants. This project is ongoing.

Silent Speech: restoring the power of speech to people whose larynx has been removed (2016-2017)

White Rose University Consortium Collaboration Fund

I was named researcher on this collaborative research project between universities of Sheffield, York and Leeds on restoring speech to laryngectomy patients, combining acoustic modelling and machine learning methods.

Improving the naturalness of statistical parametric speech synthesis using MRI-based vocal tract models (2016)

Japan Society for the Promotion of Science Summer Program Fellowship

I was selected to undertake a fellowship in Japan in summer 2016, in partnership with Tokuda and Nankaku Laboratory at Nagoya Institute of Technology in Japan. This project sought to incorporate a digital waveguide mesh vocal tract model into neural network based text-to-speech system, with intelligible speech being successfully produced by the resulting system.

External activities

Invited talks and conferences

• J. A. Gonzalez, P. D. Green, D. T. Murphy, A. J. Gully, and J. M. Gilbert, “Silent Speech: Restoring the Power of Speech to People whose Larynx has been Removed,” presented at IberSPEECH 2018, Barcelona, Spain, November 2018.
• A. J. Gully and D. T. Murphy, "Three-dimensional vocal tract modelling for forensic phonetics applications", presented at 27th Annual Conference of the International Association for Forensic Phonetics and Acoustics (IAFPA 2018), Huddersfield, UK, July-August, 2018.
• A. J. Gully, "Improving the naturalness of synthetic speech using MRI-based physical models" (poster), presented at SET for Britain, House of Commons, London, UK, March 7, 2016.
• A. J. Gully and D. M. Howard, "Perceived naturalness of a 3D dynamic digital waveguide mesh model of the vocal tract," presented at the 11th Pan-European Voice Conference (PEVOC-11), Florence, Italy, August 31 - September 2, 2015.

Publications

• A. J. Gully, P. Foulkes, J. P. French, P. T. Harrison, and V. Hughes, “The Lombard effect in MRI Noise,” in Proceedings of the International Congress of Phonetic Sciences 2019 (in press).
• A. J. Gully and B. V. Tucker, “Modeling voiced stop consonants using the 3D dynamic digital waveguide mesh vocal tract model,” in Proceedings of the International Congress of Phonetic Sciences 2019 (in press).
• A. J. Gully, H. Daffern and D. T. Murphy, "Diphthong Synthesis Using the Dynamic 3D Digital Waveguide Mesh," IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 26, no. 2, pp. 243-255, Feb. 2018.
• A. J. Gully, T. Yoshimura, D. T. Murphy, K. Hashimoto, Y. Nankaku and K. Tokuda, "Articulatory Text-to-Speech Synthesis using the Digital Waveguide Mesh driven by a Deep Neural Network," in Proc. INTERSPEECH 2017, Stockholm, Sweden, Aug. 2017