Amazon Transcribe announces a new speech foundation model-powered ASR system that expands support to over 100 languages
Blog
This article announces a new speech recognition model from Amazon Transcribe that expands support to over 100 languages. The new model uses a multi-billion parameter speech foundation model that is trained on millions of hours of unlabeled audio data across languages. It delivers significant accuracy improvements of 20-50% for most languages and 30-70% for telephony speech.
Specifically, the article covers:
- Benefits of the new model, such as improved accuracy, better punctuation and capitalization, and support for over 100 languages
- How companies like Carbyne are using Amazon Transcribe for multi-lingual emergency response
- Key features like automatic punctuation, custom vocabulary, language identification, speaker diarization, and confidence scores
- Use cases across industries like contact centers, media, and content accessibility
- How to get started with Amazon Transcribe using AWS CLI, console, and SDKs
- Example transcription output showing different views like transcripts, speaker labels, channel labels, items, and segments
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.