Service Layer medium complexity mobile
0
Dependencies
0
Dependents
0
Entities
0
Integrations

Description

On-device service responsible for audio signal preparation before it is passed to the native recognition engine. Manages microphone input levels, applies noise gating to discard ambient noise frames, and normalizes audio for higher recognition accuracy. All processing happens locally — no audio data is transmitted externally.

Feature: Speech-to-Text Input

audio-processing-service

Responsibilities

  • Normalize microphone input level for consistent recognition quality
  • Apply noise gate to filter out ambient noise during silent pauses
  • Provide real-time audio amplitude data for waveform visualization
  • Detect end-of-speech silence for automatic stop triggering

Interfaces

startCapture()
stopCapture()
Stream<double> get amplitudeStream
Stream<AudioFrame> get processedFrameStream
setNoiseGateThreshold(double threshold)
setNormalizationLevel(double level)
getAudioLevel()
dispose()