Michele LaurelliBlog

ASR (Automatic Speech Recognition)

ASR (Automatic Speech Recognition): Automatically converting speech into text through acoustic and language models. The system handles voice, speed and noise variation and can add punctuation and timestamps. WER and tests by accent or environment measure quality better than a single example.

In italiano: ASR (Automatic Speech Recognition)Legge II · Probabilità, softmax e modelli generativi

The system handles voice, speed and noise variation and can add punctuation and timestamps. WER and tests by accent or environment measure quality better than a single example.

Examples

  • Meeting transcription
  • Automatic captions
  • Voice commands