{
  "schemaVersion": "1.0",
  "updates": [],
  "companyNews": [],
  "id": "b884",
  "language": "zh",
  "name": "WhisperX",
  "originalName": "WhisperX",
  "library": "tools",
  "category": "语音识别与输入",
  "categoryKey": "语音识别与输入",
  "categoryKeys": [
    "语音识别与输入"
  ],
  "categories": [
    "语音识别与输入"
  ],
  "entryPoints": [],
  "summary": "组合批量语音识别、音素强制对齐和说话人分段，生成带词级时间戳及匿名说话人标签的转写结果。",
  "introduction": "WhisperX 将 faster-whisper 转写、VAD 分段、语言相关的音素对齐模型与 pyannote 说话人分段串联，适合为长录音制作词级时间轴、字幕或带说话人标签的文字稿。识别、对齐和说话人分段是不同步骤；当前 pyannote 模型需接受其使用协议并提供 Hugging Face 访问令牌。重叠语音、字典外词和数字等可能无法准确对齐，标签也不等同于识别真实姓名。",
  "capabilities": [],
  "company": {
    "name": "WhisperX",
    "originalName": "WhisperX"
  },
  "officialUrl": "https://github.com/m-bain/whisperX",
  "profileUpdatedOn": "2026-10-22",
  "sources": [
    {
      "kind": "product",
      "url": "https://github.com/m-bain/whisperX"
    }
  ],
  "news": [],
  "links": {
    "interactive": "products/whisperx--b884.html",
    "reading": "read/zh/products/whisperx--b884.html",
    "markdown": "read/zh/products/whisperx--b884.md",
    "json": "read/zh/products/whisperx--b884.json"
  }
}