RSDK-14075 Add wake-word-miss sensor #24
Conversation
oliviamiller
changed the title
oliviamiller
changed the title
| if yielded: | ||
| self._refresh_conversation_window() | ||
|
|
||
| async def _maybe_push_miss(self, pcm_bytes: bytes, max_score: float) -> None: |
There was a problem hiding this comment.
A little confused where this gets called?
There was a problem hiding this comment.
it gets called in _run_detection
| miss_sensor_name = str(attrs.get("wakeword_miss_sensor", "")).strip() | ||
| instance.miss_sensor = None | ||
| if miss_sensor_name: | ||
| try: |
There was a problem hiding this comment.
Should the sensor dep handling happen here or the config validation?
There was a problem hiding this comment.
moved to validation
| "wake-miss captures disabled", | ||
| miss_sensor_name, | ||
| ) | ||
| nm_thresh = attrs.get("near_miss_threshold", None) |
There was a problem hiding this comment.
Do we need to validate this threshold further? What happens if it is greater than oww_threshold?
There was a problem hiding this comment.
its it greater then no miss data would be saved, i added a check to validate to error if configured that way
| "duration_ms": duration_ms, | ||
| "audio_wav_b64": wav_b64, | ||
| } | ||
| await self.miss_sensor.do_command(cmd) |
There was a problem hiding this comment.
Any issues awaiting the cloud upload here? If this is slow does it interfere with mic streaming?
There was a problem hiding this comment.
good point, I didn't notice any issues while testing but this should run in the background to not block the audio loop, updating
2 participants
Add wakeword-miss-sensor
Summary
Testing
Configuration example added to README under the new model section.