#Is Wyoming Protocol getting further updates?

1 messages · Page 1 of 1 (latest)

valid heron
#

I have just finished building 2 Wyoming Satellites and have come to find out that perhaps the HA team has stopped working on this and switched to ESPhome. Is this the case, because I would hate to have to throw out this hardware. If they are still working on it, is there a roadmap to see when new features will hit, like continued conversation, music streaming and the like. Thank for any help anyone can give me.

tepid bolt
#

Yeah, Wyoming needs a bit love.

cold comet
#

Wyoming is still missing something very important (although I believe it's more a component of the pipeline than of Wyoming itself, it should still be supported by Wyoming if it's part of the pipelines), and that is speaker identification — for example, using Resemblyzer or SpeechBrain.

This is very important now that security is being added to Home Assistant. What’s the point of applying security to the UI but not to voice commands?

Let's hope that not only Wyoming, but the entire voice pipeline, won't be abandoned.

cold comet
#

I ran some tests with SpeechBrain on my Wyoming proxy and it works pretty well—the proxy can tell whether it’s me or my wife speaking. The problem is that this kind of feature needs to be implemented in the core; for example, voice events should include the user’s identifier.

Let’s hope to see this in Home Assistant someday.

forest dirge
#

Would also love a way to trigger the wake word from the server.

tepid bolt
forest dirge
# tepid bolt You mean start conversation?

Kind of. What I would like is a way to manually trigger the satellite to start listening from a button push on the device, like the one I have here, or a dashboard. Like a manual wake button.

I’ve looked at start_conversation, but it doesn’t recognize any of my devices, and isn’t entirely what I am looking for.

IMG_3288.jpg
vale tiger
#

If your satellite is ESPHome based you can call voice_assistant.start

cold comet
#

First of all, sorry for continuing off-topic.

You can add a service directly in your satellite's YAML. For VPE, for example, it's enough to add this in the API section:

`api:
services:

  • service: start_assist
    then:
    • if:
      condition:
      switch.is_off: master_mute_switch
      then:
      # If a timer is ringing: Stop it, do not start the voice assistant (We can stop timer from voice!)
      - if:
      condition:
      switch.is_on: timer_ringing
      then:
      - switch.turn_off: timer_ringing
      # Stop voice assistant if running
      else:
      - if:
      condition:
      voice_assistant.is_running:
      then:
      voice_assistant.stop:
      # Stop any other media player announcement
      else:
      - if:
      condition:
      media_player.is_announcing:
      then:
      - media_player.stop:
      announcement: true
      # Start the voice assistant and play the wake sound, if enabled
      else:
      - if:
      condition:
      switch.is_on: wake_sound
      then:
      - script.execute:
      id: play_sound
      priority: true
      sound_file: !lambda return id(wake_word_triggered_sound);
      - delay: 300ms
      - voice_assistant.start:`
#

In case you're using a different YAML, the logic is the same: add a service that replicates your satellite's logic. I think with an Athom M5 with the firmware with microwakeword it would be something like this:

`api:
services:

  • service: start_assist
    then: on_wake_word_detected:
    • voice_assistant.start:
      wake_word: !lambda return wake_word;`
tepid bolt
forest dirge
wooden sluice
forest dirge
forest dirge
#

Eh. Unfortunately this does not seem stable - at least with my setup. It triggered after about a minute delay and then never stopped listening. Will reach out to the dev

valid heron
#

Alright this seems like the best place to ask this then.
Is there anyway to have continued conversation by just telling the satellite to keep listening for a few seconds and responding? Should say that I'm using Wyoming not esp.

wooden sluice