WAMI: Web-Accessible Multimodal Interfaces
WAMI is a simple way to add speech recognition capabilities to any web page. The API was developed by the Spoken Language Systems group at the MIT Computer Science and Artificial Intelligence Laboratory.
The students who developed WAMI have graduated and are no longer able to support and maintain it. We will keep WAMI running as resources permit, but cannot provide additional support.
Due to hardware maintenance, WAMI is likely to be unavailable for a while starting on August 5, 2013.
- Grammar Sandbox. Try out the service with this example.
- WAMI 2.0 Documentation. Get started devloping a speech-enabled site.
- Developers Portal. Create and manage your developer account.
If you just want recording capabilites, but can't wait for HTML5, try this:
- The WAMI recorder is a light-weight Flash client for recording audio.
Try out some of these WAMI examples:
- Play a game of Voice Tic-Tac-Toe
- Do some math on the Voice Calculator
- The Parrot will repeat a few simple phrases.
- MIT Map speech interface to the MIT campus map.
- We also have a recognizer for Mandarin Chinese. Here are two games that teach language learners driving directions and dates and times.
Please note: all audio sent to MIT's servers will be logged for research purposes. Your WAMI application should notify users of this fact.