In this demonstration paper, we introduce the first publicly available Speech-To-Text transcription service for the Latvian language. We present its main features, the details of automatic speech recognition (ASR) system used in this service, software architecture, and an evaluation of recognition quality. The service will provide regular people with the opportunity to transcribe their own audio files for various purposes, such as lectures, meetings, etc. Also, the users will be given an opportunity to give their evaluation and feedback about the quality and usability of this service, which will be used by developers to make changes in the ASR in order to improve it.
Bibliographic reference. Salimbajevs, Askars / Strigins, Jevgenijs (2015): "Latvian speech-to-text transcription service", In INTERSPEECH-2015, 722-723.