mirror of
https://github.com/R0m1k3/EveFlow.git
synced 2026-10-11 17:29:03 +02:00
Claude
0d284fea9e
- Catalog: 3.3 MB zipformer keyword-spotting model (kws-en). - shared/keywords: BPE encoding of wake phrases (SentencePiece table for common words, greedy longest-match fallback over the model vocabulary) and keywords file builder. - Worker: KeywordSpotter stream fed with 16-bit PCM, detections pushed as unsolicited messages; engine derives the keywords file, maps sensitivity to threshold/score, forwards detections to the renderer and re-arms after a worker restart. - Renderer: WakeListener keeps one microphone stream, batches 256 ms frames to the spotter and captures the command on the same stream after detection (pre-roll, VAD), then resumes spotting; the wake word also interrupts speech. Settings: wake mode (off / always-on / transcript filter), keyword, sensitivity, status and one-click model download; HUD caption shows the active keyword. - Validated: detection in the worker (fork) and through the real Electron IPC path on Kokoro audio, no false positive on an English recording; 23 unit tests. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017Wn5VX9HNbJ7N54hR24u9Y
108 lines
2.4 KiB
TypeScript
108 lines
2.4 KiB
TypeScript
/** Local voice engine contract (sherpa-onnx in a utility process). Shared by main and renderer. */
|
|
|
|
export type VoiceModelKind = 'stt' | 'tts' | 'kws';
|
|
export type VoiceEngineKind = 'whisper' | 'sense-voice' | 'nemo-transducer' | 'kokoro' | 'piper' | 'kws-transducer';
|
|
|
|
export interface VoiceSpeaker {
|
|
id: number;
|
|
name: string;
|
|
lang: string;
|
|
}
|
|
|
|
export interface VoiceModelSpec {
|
|
id: string;
|
|
kind: VoiceModelKind;
|
|
engine: VoiceEngineKind;
|
|
name: string;
|
|
description: string;
|
|
languages: string[];
|
|
sizeMb: number;
|
|
url: string;
|
|
/** Folder created by the archive (files are referenced relative to it). */
|
|
dir: string;
|
|
/** Files that must exist once installed. */
|
|
files: string[];
|
|
speakers?: VoiceSpeaker[];
|
|
sampleRate?: number;
|
|
recommended?: boolean;
|
|
}
|
|
|
|
export interface VoiceModelStatus extends VoiceModelSpec {
|
|
installed: boolean;
|
|
downloading: boolean;
|
|
installedBytes: number;
|
|
}
|
|
|
|
export interface VoiceDownloadProgress {
|
|
id: string;
|
|
phase: 'download' | 'extract' | 'done' | 'error' | 'cancelled';
|
|
received: number;
|
|
total: number;
|
|
percent: number;
|
|
message?: string;
|
|
}
|
|
|
|
export interface TranscribeRequest {
|
|
modelId: string;
|
|
wav: Uint8Array;
|
|
language: string; // 'fr', 'en', 'auto'
|
|
}
|
|
|
|
export interface TranscribeResult {
|
|
text: string;
|
|
language?: string;
|
|
durationMs: number;
|
|
audioSec: number;
|
|
}
|
|
|
|
export interface SynthesizeRequest {
|
|
modelId: string;
|
|
text: string;
|
|
speaker: number;
|
|
speed: number;
|
|
}
|
|
|
|
export interface SynthesizeResult {
|
|
wav: Uint8Array;
|
|
sampleRate: number;
|
|
durationMs: number;
|
|
audioSec: number;
|
|
}
|
|
|
|
export interface KwsStartRequest {
|
|
modelId: string;
|
|
/** Wake phrases in plain text (e.g. "jarvis", "hey jarvis"). */
|
|
keywords: string[];
|
|
/** 1 (strict) .. 5 (eager) */
|
|
sensitivity: number;
|
|
}
|
|
|
|
export interface KwsDetection {
|
|
keyword: string;
|
|
at: number;
|
|
}
|
|
|
|
export interface VoiceEngineStatus {
|
|
available: boolean;
|
|
error?: string;
|
|
version?: string;
|
|
loaded: string[];
|
|
modelsDir: string;
|
|
}
|
|
|
|
export const VOICE_IPC = {
|
|
status: 'voice:status',
|
|
modelsList: 'voice:models:list',
|
|
modelsDownload: 'voice:models:download',
|
|
modelsCancel: 'voice:models:cancel',
|
|
modelsRemove: 'voice:models:remove',
|
|
modelsProgress: 'voice:models:progress',
|
|
transcribe: 'voice:transcribe',
|
|
synthesize: 'voice:synthesize',
|
|
unload: 'voice:unload',
|
|
kwsStart: 'voice:kws:start',
|
|
kwsStop: 'voice:kws:stop',
|
|
kwsAudio: 'voice:kws:audio',
|
|
kwsDetected: 'voice:kws:detected'
|
|
} as const;
|