The mngu0 articulatory corpus
Welcome to the mngu0 multimodal, single-speaker speech articulation corpus.
This is a corpus of articulatory data of different forms (EMA, MRI, video, 3D scans of upper/lower jaw, audio etc.) acquired from one male British English speaker. The purpose of this page is to:
- distribute the data (in a variety of forms and parameterisations); and
- identify and keep track of research work which makes use of this data
The first part of this corpus to be released is a large set of utterances recorded by the male British English speaker in a Carstens AG500 Electromagnetic Articulograph. More details can be found in the initial announcement paper for mngu0, which appeared at the Interspeech 2011 conference.
Download mngu0 data
First, you will need to register to obtain a unique UserID. Send me an email with the following details to request your UserID:
- full name and email address
- country and organisation (e.g. University of Edinburgh, or Google, or your company etc.)
- work sector (academic research, academic teaching, education, commercial research, etc.)
- intended use (what would you like to use mngu0 for? Speech synthesis/recognition research? Articulatory animation? Something else?)
I'll reply to send your personal UserID (done manually, so there may be some delay). You can then use that to download whatever individual files you want below. When you click the "submit" button, you will receive an email with a link to the file you have selected.