家电科技 ›› 2019, Vol. 0 ›› Issue (4): 54-58.doi: 10.19784/j.cnki.issn1672-0172.2019.04.004

• 论文-制冷技术 • 上一篇    下一篇

基于深度学习的语音识别技术研究及其在空调上的应用

毛跃辉   

  1. 珠海格力电器股份有限公司 广东珠海 519070
  • 出版日期:2019-08-01 发布日期:2019-08-19

Research on speech recognition technology based on deep learning and its application in air conditioning

MAO Yuehui   

  1. Gree Electric Appliances Inc. of Zhuhai Zhuhai 519070
  • Online:2019-08-01 Published:2019-08-19

摘要: 针对人工智能领域开展语音识别技术研究及空调应用开发,采用麦克风阵列技术,通过独有的低功耗唤醒和识别双引擎技术、基于混合深度学习模型的语音识别算法、融合自适应远场降噪技术等,抑制环境噪声并进行人声增益、声源定位和波速形成,实现高准确率远场语音识别和强可靠性运作。

关键词: 语音识别, 双引擎模式, 自适应降噪, 空调误识别抑制, 空调应用

Abstract: In view of the research and application development of speech recognition technology in the field of artificial intelligence, dual microphone array technology is adopted to suppress environmental noise and carry out human voice through unique low power wake-up and recognition dual engine technology, speech recognition algorithm based on mixed depth learning model, fusion adaptive far field noise reduction technique and so on. Gain, acoustic source localization and wave velocity formation are used to achieve high accuracy far-field speech recognition and strong reliability operation.

Key words: Speech recognition, Dual engine mode, Adaptive noise reduction, Air conditioning false identification suppression, Air conditioning applications