The Emotion in Speech Project was a joint research project between the Speech Laboratory at the University of Reading and the Department of Psychology at the University of Leeds, and was funded by the Ministry of Defence (DERA, Malvern) and ESRC (Award No. R000235285).
The project is now complete and copies of the corpus have been deposited with the ESRC Data Archive.
Of the many types of suprasegmental and affective information that have been found to occur in speech, relatively few have been coded in such a way as to permit inclusion of them in large-scale machine-readable speech databases. However, as the demand for more natural and unconstrained speech material grows, it becomes increasingly necessary to look at ways of doing this.
This project brought together expertise in phonetics and phonology and in cognitive psychology in order to examine emotional speech and to produce a database of such speech to put alongside the emotionally neutral material found in most spoken language databases.
As a first step it is necessary to devise suitable coding systems. For the prosodic and paralinguistic coding we have based our analysis on the ToBI system (Silverman et al, 1992). This system is very valuable for the annotation of corpus material, but is in our opinion too phonologically-oriented to permit detailed representation of the phonetic and paralinguistic information which we consider relevant to analysis of emotional speech. However, the "Miscellaneous" tier of this system does allow a considerable amount of information to be added to supplement the Tonal and Break Index tiers, and we have chosen as the basis for doing this the prosodic and paralinguistic feature system devised by Crystal (1969), though with significant alterations that we have felt it necessary to introduce. We have an example of the type of data that the project generated.
Secondly, it is necessary to go beyond the very impoverished conceptual and terminological framework that has been used by phoneticians in describing the attitudes and emotions transmitted by vocal means, and use a classification system that reflects current thinking on emotion by cognitive psychologists. An additional tier has been added to the ToBI tier system for representing emotions in a way that links them temporally to the text and the acoustic waveform.
Papers resulting from the project.
The only analysis to date to take into account the various emotions annotated in the data has been carried out by Richard Stibbard for his PhD thesis.
Reading University (speech variables):
Leeds University (emotional variables):
We are sorry but because of copyright restrictions the Emotion in Speech Corpus is not available for distribution.
Professor Peter Roach (llsroach@reading.ac.uk)
September 2000