Training Trains Internship
Student login
← Natural Language Processing learning path
Natural Language Processing · Module 9 of 10

Speech and multilingual NLP

Work with speech and many languages.

📖 Key ideas

  1. Speech-to-text (ASR) converts audio to text (Whisper).
  2. Text-to-speech (TTS) generates voice.
  3. Multilingual models handle many languages in one model.
  4. Indian language resources: AI4Bharat, Bhashini.
  5. Transliteration converts scripts (Hindi in Latin letters).
💡 Real-life example: ASR is a stenographer typing every word spoken in a meeting.
ASRAutomatic Speech Recognition
TTSText-to-Speech
WhisperOpenAI speech recognition model
TransliterationConverting between scripts