Class Diagram
This diagram illustrates the core architecture of AAC Voice API, showing the relationships between main classes and their responsibilities
Figure 2 Class Architecture of our API
index
This is the main entry point for npm package 'aac-voice-api'.
It exports the core classes, functions, and types that
make up the public API of the library.
AACVoiceApi
The main entry point for the API. This will be a facade and the only class that a developer will have to initialize. Initializes voice command recognition and manages the lifecycle of voice listening sessions.
SpeechConverter
Handles real-time speech-to-text conversion using the Whisper model. It manages audio input, preprocessing, and transcription directly in the browser.
CommandHistory
Keeps track of a chronological log of commands.
commandLibrary
Contains a HashMap that:
- Can be called by CommandMapper.
- Maps a String command to the corresponding GameCommand.
AudioInputHandler
Handles raw audio stream processing from the user's microphone. Manages Web Audio API components and converts audio into processable chunks.
showHistoryPopup
Opens a modal that shows the latest commands and auto-refreshes every second. Storage is unbounded; UI renders only a window (latest PAGE items) for speed.
SpeechToText
Consumes (from AudioHandlerInput) and converts audio chunks into text transcriptions for command recognition.
CommandMapper
Maps recognized speech text to configured voice commands and triggers the appropriate functions
SpeechSeparation
Processes audio to separate speech from background noise, improving recognition accuracy
Key Relationships
- AACVoiceApi serves as the orchestrator, containing and coordinating all major components
- AudioHandlerInput captures and preprocessors audio before passing it to the speech recognition pipeline