- AWS’s transcription platform can now recognise 100 different spoken languages after an AI update added 21.
- The update mainly includes accuracy improvements, readability enhancements and the introduction of new AI features.
- This update is set to boost AWS popularity, competitiveness, and offer users a more convenient transcription experience.
Amazon Transcribe, AWS’s transcription platform, can now recognize over 100 spoken languages.
Announced at the recent AWS re: Invent event, the update adds 21 new languages to its existing 79 languages, bringing the total to 100. The new languages include Afrikaans, Albanian, Amharic, Armenian, Azerbaijani, Bengali, Bosnian, Bulgarian, Burmese, Croatian, Dari, Estonian, Georgian, Hausa, Kannada, Khmer, Kurdish, Latvian, Lithuanian, Macedonian, Malayalam, Marathi, Mongolian, Nepali, Pashto, Persian, Serbian, Sinhala, Slovak, Slovenian, and Tamil.
Accuracy improvements
Amazon Transcribe utilises sophisticated self-supervised algorithms trained on vast amounts of unlabelled audio data to grasp universal speech patterns across various languages and accents. Through smart data sampling, it ensures a high level of accuracy, particularly in languages that lack representation.
Amazon Transcribe has significantly enhanced accuracy through its speech base model, boasting improvements of 20-50% for most languages. Telephone speech recognition sees even more substantial gains, ranging from 30-70%.
The software has implications for verious industries, including, for example, emergency response units.
According to Alex Dizengof, Co-Founder and CTO of Carbyne, a software company that develops cloud-based, mission-critical contact center solutions for emergency call responders, “By leveraging Amazon Transcribe’s new multilingual foundation model powered ASR, Carbyne will be even better equipped to democratize life-saving emergency services, because Every. Person. Counts.”
Also read: Is Formula One’s use of AI for track limits enforcement better than using humans?
Readability enhancements
As well as the large directory of supported languages, Amazon Transcribe has improved readability and the user experience. This expansive ASR (Automatic Speech Recognition) model can deliver a seamless speech recognition experience for both existing and new customers in batch mode—no modifications to API end nodes or input parameters required.
New AI features
New AI features include automatic punctuation, custom vocabulary, automatic language recognition, and custom vocabulary filtering, thus increasing the flexibility and customisation of the system.
In addition, the system’s extended support for different accents, noise environments, and acoustic conditions helps generate more accurate output, enabling users to integrate speech technology more effectively into applications.
Overall, this update provides users with an easier and more flexible experience, and Amazon Transcribe’s update is expected to make it widely popular and more competitive.
Also read: Is OpenAI’s rush for revenue a threat to humankind?
