2. Modern speech synthesis: communication aid personalisation Sarah Creer Stuart Cunningham Phil Green Clinical Applications of Speech Technology University of Sheffield

3. Introduction Building voices for VIVOCA (communication aids) Speech synthesis techniques Future research: personalisation of synthetic voices

4. Current speech synthesis: communication aids High quality voices available E.g. Toby Churchill Lightwriter – DECtalk™ (Fonix) for American English – Acapela for British English Personalisation limited: age, gender, language

5. Personalisation Voice = identity – Gender – Age – Geographic background – Socio-economic background – Ethnic background – As that individual Maintains social relationships Maintains social closeness Sets group membership

6. VIVOCA Voice Input Voice Output Communication Aid Speech Recognition Dysarthric speech input Text-to-Speech Synthesis Recognised text Intelligible synthesised speech output Retain elements of clients’ identity for synthesised speech output Intelligible and personalised synthesised speech output

7. VIVOCA: personalisation Sheffield/Barnsley user group Retain local accent – geographic identity Speaker database – Arctic database: 593 + 20 sentences Professional local speakers – Ian McMillan – Christa Ackroyd

8. Concatenative synthesis Input data Text input Synthesised speech Speech recordings Unit segmentation Unit database Unit selection Concatenation + smoothing i a sh Festvox: http://festvox.org/ +… ++ …

9. Concatenative synthesis High quality Natural sounding Sounds like original speaker  Need a lot of data (~600 sentences)  Can be inconsistent  Difficult to manipulate prosody

10. HMM synthesis yes yes

11. HMM synthesis procedure Input data Text input Speaker model Synthesised speech Speech recordings TrainingSynthesis e t HTS http://hts.sp.nitech.ac.jp/

12. HMM synthesis Consistent Intelligible Needs relatively little input (~20 mins) Can be adapted with small amount of data (>5 sentences) Easier to manipulate  Buzzy quality  Less natural than concatenative

13. Future research Further personalisation for individuals with progressive speech disorders – Capturing the essence of a voice Voice banking – Before deterioration Adaptation using HMM synthesis – Before or during deterioration

14. Thank you This work is sponsored by EPSRC Doctoral Training grant For further details of VIVOCA see: http://www.shef.ac.uk/cast/ Email: S.Creer@dcs.shef.ac.uk