Bangla Speech to Text
The SUST Bangla Language Processing research team is working on Bangla Speech Processing (Bangla Text to Speech – TTS or Speech synthesis, Bangla Speech to Text – STT or Automatic Speech Recognition - ASR) etc. In the Bangla Speech to Text research area – the main goal was to develop a prototype of Bangla ASR system for Bangla speech recognition on mobile app and web app. The SUST had developed and realized the alpha version of the system on the end of the last September 2018. Now we have realized the Beta version of the ASR web app. The core technology - LSTM-RNN CTC have been used to do the Acoustic Modeling (AM) and 250 hours of speech corpus have been used for training AM of this ASR system. The systems are available as web apps and will be also available as mobile apps.