Voice acknowledge words, such as "Hmm?", after wake word is used


So I thought it would be a nice idea to have acknowledgement words.

When I talk to Siri, “Hey Siri”, she replies, “Hmm?” - as an indication that she is listening for a command.
“Hmm?” In this case would be the acknowledgement word.

It would be pretty cool to have a selectable list, or randomly shuffled acknowledgement words, such as

  • Yes?
  • Hmm?
  • I’m listening
  • etc.

Which would be played after openwakeword detects the wake word.

I concur that assist needs some sort of chime or acknowledgement that it is listening.
I have a pi setup with openwakeword and Wyoming and would like to ultimately phase out my google homes … but not having any indication that it is listening is a deal breaker for the family.

You both mentioned OpenWakeWord, which means you’re running a RasPi with wyoming-satelite ?

The “–awake-sound” option plays the chime (or whatever wav file) to indicate it’s listening, and “–done-sound” to indicate it has finished listening. Refer to GitHub - rhasspy/wyoming-satellite: Remote voice satellite using Wyoming protocol

script/run --host --token eyJhbG...VKQwgsCDQ 
    --mic-device plughw:CARD=ABTWPDQ0222M,DEV=0  --awake-sound sounds/awake.wav
    --snd-device plughw:CARD=Headphones,DEV=0  --done-sound sounds/done.wav 
    --auto-gain 5 --vad webrtcvad  --wake-word wyoming

You could record your own wav files, even using Piper to get the same voice shouldn’t be too difficult.

I have voted for this because I like the idea of it randomly selecting a response.

Or an ESP.

That said, with the ESP’s you include a TTS message OR sound clip playback etc after:

  id: va
  microphone: box_mic
  speaker: box_speaker
  use_wake_word: true
  noise_suppression_level: 2
  auto_gain: 31dBFS
  volume_multiplier: 2.0
  vad_threshold: 3

This will only work with newer firmware where the device is capable of being a media_player as well as voice assistant. eg.

1 Like

+1 for this.

Now that we have wake words, LED indication is fine but useless unless you’re line of sight.

Really want an option to give a verbal/auditory acknowledgement of wake word detection via the attached speaker of the M5 stack echo or an external media player

@itnassol this may be more what you were looking for.

This would be a nice feature. Would be cool to set the sound it makes or have a set of sounds that are played randomly.