PALS0039 Introduction to Deep Learning for Speech and Language Processing
SPEECH, HEARING & PHONETIC SCIENCES
UCL Division of Psychology and Language Sciences

Available Datasets

This document describes the datasets that have been prepared for demonstrations of deep learning on the course.


Vowel data

List of vowel formant frequencies for English monophthongs by multiple speakers described by gender and height.


Age

Processed speech recordings from the Accents of the British Isles corpus. Columns identify speaker, gender, accent and age. Each recording is analysed using the OpenSMILE toolkit, into 6373 features.


Word count: . Last modified: 22:45 11-Mar-2022.