EMIME Foregrounds

As deliverables of the FP7 EMIME project, several tools and databases have been released and become publicly available. My HTS voice building scripts (used for building the CSTR/EMIME HTS 2010 system) that support parallel processing and several average voice models (English, Finnish and Mandarin) for HMM-based speech synthesis have also been released from the site and been available for download. The use of the pre-trained average voice models are restricted to research use.

Download URL

Voice building tools for CSTR/EMIME HTS 2010 system: https://www.emime.org/participate/tools
English, Finnish and Mandarin average voice models: http://www.emime.org/participate/tts-average-voice-models