AudioSendModeRequest is not exposed to the API

## Issue description

There is this [AudioSendModeRequest](https://github.com/digital-dream-labs/vector-cloud/blob/main/internal/proto/external_interface/messages.pb.go#L7434-L7439) method allowing to set the [AudioProcessingMode](https://github.com/digital-dream-labs/vector-cloud/blob/main/internal/proto/external_interface/messages.pb.go#L393-L416) to other value than the [default value](https://github.com/digital-dream-labs/vector-cloud/blob/main/internal/proto/external_interface/external_interface.pb.go#L217)
of AudioProcessingMode_AUDIO_UNKNOWN.

Unfortunatelly, it seems that there is no exposed RPC call to invoke that request from API to set up the AudioProcessingMode in vic-gateway to other value than this initial value, and therefore the only sound currently received from the AudioFeed is constant test tone...

## Suggested Resolution
- to expose this call as rpc to enable the actual audio to be streamed through AudioFeed.

## Goal
- to use this for what it was originally intended - to set it up for example to AUDIO_VOICE_DETECT_MODE, and process the audio with custom voice recognition systems even externally with the API, for all robots out there.