IndicConformer: Speech Recognition for Indian Languages 🎙️➡️📜

This Gradio demo showcases IndicConformer, a speech recognition model for 22 Indian languages. The model operates in two modes: CTC (Connectionist Temporal Classification) and RNNT (Recurrent Neural Network Transducer), providing robust and accurate transcriptions across diverse linguistic and acoustic conditions.

How to Use:

  1. Upload or record an audio clip in any supported Indian language.
  2. Select the mode (CTC or RNNT) for transcription.
  3. Click "Transcribe" to generate the corresponding text in the target language.
  4. View or copy the output for further use.

🚀 Try it out and experience seamless speech recognition for Indian languages!
Target language
Examples
Input speech Target language