An Automatic Speech Transcription System for Manipuri Language

Tanvina Patel, Krishna D N, Noor Fathima, Nisar Shah, Mahima C, Deepak Kumar, Anuroop Iyengar


Development of speech technologies in Indian languages has witnessed a steep improvement recently. In this work, we present our efforts in building various speech technology applications for Manipuri language. For the language at hand, we initially perform Language identification (LID) task. This is followed by speech-to-text (STT) and Keyword Search (KWS). In addition, we build a Speaker Diarization (SD) framework as well. The speech modules are integrated together to extract information from the speech signal. Currently, the platform is build for Manipuri and English language and can be extended to other languages as well. A visual User Interface (UI) is available for demonstration purpose where given a set of speech files the services from all the mentioned speech modules can be used.


Cite as: Patel, T., N, K.D., Fathima, N., Shah, N., C, M., Kumar, D., Iyengar, A. (2018) An Automatic Speech Transcription System for Manipuri Language. Proc. Interspeech 2018, 2388-2389.


@inproceedings{Patel2018,
  author={Tanvina Patel and Krishna D N and Noor Fathima and Nisar Shah and Mahima C and Deepak Kumar and Anuroop Iyengar},
  title={An Automatic Speech Transcription System for Manipuri Language},
  year=2018,
  booktitle={Proc. Interspeech 2018},
  pages={2388--2389}
}