Skip to content
This repository was archived by the owner on May 6, 2022. It is now read-only.

Release 3.2.0

Choose a tag to compare

View all tags
@timmywil timmywil released this 26 Apr 19:27
· 34 commits to develop since this release
3.2.0
e4e3f1e
This commit was signed with the committer’s verified signature. The key has expired.
timmywil Timmy Willison
GPG key ID: B1D1D9907F95E829
Expired
Verified
Learn about vigilant mode.

3.2.0 (2021-04-26)

Features

This release adds experimental support for the full speech pipeline, with support for wake word, keyword, and voice activity detection. The speech pipeline uses tensorflow.js off the main thread in a web worker to process speech using Spokestack models.

Note: The speech pipeline is currently only supported in Blink browsers such as Chrome, Edge, Opera, Brave, and Vivaldi. It uses advanced APIs not yet available in the other browsers (specifically, some webgl APIs used in tensorflow.js).

Bug Fixes

  • ASR one-off requests (using the asr function) reported the first match rather than the final match. This bug did not affect websocket ASR.
  • Added stricter type checking for the sampleRate argument on ASR requests
  • Improved error reporting
Assets 2
Loading