Improving transcription accuracy - Amazon Transcribe
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Improving transcription accuracy

If your media contains domain-specific or non-standard terms, such as brand names, acronyms, technical words, and jargon, Amazon Transcribe might not correctly capture these terms in your transcription output.

To correct transcription inaccuracies and customize your output for your specific use case, you can create Custom vocabularies.

  • Custom vocabularies are designed to tune and boost both the recognition and the formatting of specific words in all contexts. This involves supplying Amazon Transcribe with words and, optionally, pronunciation and display forms.

    If Amazon Transcribe is not correctly rendering specific terms in your transcripts, you can create a custom vocabulary file that tells Amazon Transcribe how you want these terms displayed. This word-specific approach is most appropriate for correcting terms like brand names and acronyms.

Tip

To learn more, including how to create a custom vocabulary, refer to the Custom vocabularies section.