ENRICH’s research was done by our 14 Early Stage Researchers. Find out about their project aims below.
A computational metric for listening effort
Develop and validate metrics for listening effort, and apply them in the generation of speech enriched to lower its cognitive processing burden.
Enriching disordered speech
Design of data- or knowledge-driven voice/speech transformations to enable communication in the context of disordered speech. Speech is to be enriched by boosting its intelligibility and pleasantness.
Impact of acoustic enrichment on ease of processing
Investigate how speech perception in challenging situations impacts on cognitive load and physical stress. Examine effects of speaker age, gender, language, hearing.
Impact of linguistic enrichment on the perception of highly-degraded speech
Investigate effects on discourse comprehension of visual clarity, phonetic composition, syntactic structure, lexical familiarity, degree of predictability and word meaning.
Embedding enhancement information in the speech signal
Discover what additional information would assist enhancement algorithms running on hearing aids. Use the speech signal to carry this overtly or covertly (audio watermarking).
Synthetic speech optimised for low cognitive load
Investigate cognitive load imposed by synthetic speech, which is 'hard to listen to'. Create novel forms of synthetic speech that impose the lowest possible cognitive load.
Individual differences in enriching one's speech
Quantify cognitive speaking effort, discover why speech enrichment is easier for some people, and measure effectiveness of different strategies.
Sounding non-native: Speaking clearly in a second language
Discover how non-native speakers enrich their speech, the effort needed to do this, and the efficiency of different strategies for different listener groups.
Real-life speech intelligibility enhancement
Design new approaches for enhancing the intelligibility of distorted speech, under loudness constraints.
From conversational to clear speech
Devise parametric mappings of temporal envelope and fine structure modulations from conversational to clear speech.
Enhancement of speech emotion recognition
Develop visual materials (gestures, face models, body movements) to enrich speech emotion information, for training hearing-impaired children and adults in better emotion identification.
Top-down enrichment of speech perception
Investigate positive effects of musical training on speech intelligibility and cognitive effort of speech recognition, then determine if this applies to hearing-impaired children.
Visual enrichment for improving acoustic communication with hearing aids
Improve hearing-aids by multimodal analysis of the acoustic environment and the user’s head- and eye-movements to control the signal processing and provide additional visual cues to the user.
Assessing and reducing cognitive load of listening to speech in adverse conditions
Compare and optimise different methods of speech enhancement that might reduce cognitive load and listening effort.