search

espressif / esp-adf

Espressif Audio Development Framework
Other
1.5k stars 670 forks source link

audio_recorder: expose afe_fetch_result_t (AUD-4895) #1076

Closed stintel closed 10 months ago

stintel commented 10 months ago

Is your feature request related to a problem? Please describe.

For Willow, we're working on a feature to avoid multiple voice assistants in earshot of each other all responding to the same voice command. The feature is explained in more detail here.

The first step to make this possible has already landed in esp-sr: espressif/esp-sr#79.

Describe the solution you'd like

We would like to be able to access the wake volume from esp-sr in the Willow code.

Describe alternatives you've considered

NA

Additional context

We managed to make a working PoC, by modifying some ESP-ADF code. These changes can be found here. We will gladly open a PR to implement this, but we see our current changes to ESP-ADF as a quick hack, and would like to ask for your input before we open a PR.

Here's a video demonstrating the feature.

jason-mao commented 10 months ago

@stintel Yes, the wake volume is useful, we plan to make a break changes to support this feature and improve the compatibility.