Home Downloads

Downloads

All the resources that were gathered during the Lwazi project are freely available for you to use in creating you own automatic speech recognition or text-to-speech systems.

Data resources		Tools		Lwazi service resources
Download Lwazi data resources: Text-to-Speech (TTS) All audio recordings used for the creation of Lwazi Text-to-speech systems. Automatic Speech Recognition (ASR) Complete audio recordings and orthographic transcriptions used for Lwazi speech recognition systems. Pronunciation Dictionaries Complete Dictionaries including phoneme sound files, default-and-refine rules and aligned dictionary file. Download a phoneset for all the languages: Phoneset [PDF file, 95K] For all the previous versions of released data, please contact us		The following tools are available for download: Speect A tool that can be used for the rapid development of text-to-speech (TTS) voices in any language. ASR-Builder A tool that can be used for the construction of automatic speech recognition (ASR) systems in any language. DictionaryMaker A tool that can be used by researchers to build dictionaries in any language. Lwazi Telephony Platform A tool that can be used to easily build low-cost telephony systems using open-source software.		Download Lwazi service resources: Pilot implementation software An explanation of the software modules needed to set up the Lwazi service can be found here. Lwazi service voice prompts Download voice prompts for the Lwazi service.