New features
- set default synthesis voices to neural voice (#376)
- Use AudioWorkletNode by default for mic input (#397)
- getActivationPhrasesAsync API added to VoiceProfileClient class for receiving a list of valid activation phrases in speaker recognition enrollment phase for independent recognition scenarios.
- VoiceProfileClient's enrollProfileAsync API is now async awaitable. See this independent identification code for example usage.
Bug fixes
- fix viseme and bookmark tests failure (#393 )
- fix word level json offsets (#396 )
- fix push stream not being reconnected to with start-stop-start continuous recognition, add test (#399)
- Browser microphone audio processing now uses AudioWorkletNode instead of deprecated ScriptProcessorNode (#391)
- Correctly keep conversations alive during long running conversation translation scenarios. (#389)
- Fixed issue with recognizer reconnecting to a mediastream in continuous recognition. (#385)
- Fixed issue with recognizer reconnecting to a pushStream in continuous recognition. (#399)
- Correct word level offset calculation in detailed recognition results. (#394)