构音障碍语音识别

帮助脑卒中患者将含糊不清的语音翻译为标准中文

    Group 30
    Device frame
    语音识别

    核心功能

    Integrated digital wallet

    基于 wav2vec2 中文语音预训练模型,提取 1024 维语音嵌入特征,余弦相似度精准匹配ts. Manage and store your funds with ease, all within a single, convenient platform.
    Wallet

    语音录入与识别

    Effortlessly make and receive payments with just a few taps, ensuring quick and secure transactions every time.
    发送

    Regular monthly and weekly summaries

    Receive comprehensive monthly and weekly summaries that provide detailed insights into your financial activities, helping you track spending patterns and make informed decisions with ease.
    Report

    256-bit security encryption

    Enjoy top-tier protection with 256-bit encryption, safeguarding your data with advanced security algorithms.
    Safety

    语音嵌入模型

    模型本地运行,患者语音数据安全存储,保护个人隐私信息。
    Cart

    上传语音 智能识别

    上传患者语音或视频文件,系统自动提取音频特征进行识别

    • Group 30
      Device frame
  • Check
    Make payments

    点击选择文件,系统自动完成音频提取和特征匹配,几秒内显示识别结果

  • Check
    Solicit payments

    录入患者不同词汇的发音基线,多次录入自动计算平均嵌入向量,持续提升识别准确率

  • 建立患者专属语音基线

    为脑卒中构音障碍患者提供个性化语音识别服务,将含糊不清的发音翻译为清晰中文ve your financial goals effortlessly

    • Group 29
      Device frame
  • Check
    Monthly & weekly reports

    支持多次录入同词基线,系统自动计算平均嵌入向量,持续提升识别准确率both monthly and weekly

  • Check
    Detailed transaction records

    查看所有识别历史记录,清晰了解每一次语音翻译的结果和准确率变化趋势ew of your spending and income history

  • 专为脑卒中构音障碍患者设计

    基于 wav2vec2 中文预训练模型,采用余弦相似度匹配算法,为每位患者建立专属语音基线库ted by a vast community

    • Group 31
      Device frame
  • 200M

    基线词汇

  • 4.80+

    识别准确率

  • 30M

    活跃用户

  • 50+

    系统集成

  • 开始建立患者语音基线,体验智能识别

    选择文件即可自动识别,基于深度学习模型几秒内给出翻译结果

      Group 29
      Device frame