首页> 外国专利> APPARATUS FOR DETERMINING STATE BASED ON DEEP NEURAL NETWORK ACOUSTIC MODEL FOR VOICE RECOGNITION AND METHOD THEREOF

APPARATUS FOR DETERMINING STATE BASED ON DEEP NEURAL NETWORK ACOUSTIC MODEL FOR VOICE RECOGNITION AND METHOD THEREOF

机译:基于深度神经网络语音识别模型的状态确定装置及其方法

摘要

The present invention relates to a method for determining a state based on a deep neural network for an acoustic model for voice recognition. The method comprises: a state determining step of determining a state of voice data using a coupling coefficient and bias to an output node used in a neural network model for training voice data; and a learning step of learning the voice data using a state determining input value determined through the state determining step and the coupling coefficient and bias to the output node used in the neural network model.;COPYRIGHT KIPO 2019
机译:本发明涉及一种基于深度神经网络的用于语音识别的声学模型的确定状态的方法。该方法包括:状态确定步骤,使用耦合系数和对神经网络模型中用于训练语音数据的输出节点的偏置来确定语音数据的状态; COPYRIGHT KIPO 2019;以及通过使用状态确定步骤确定的状态确定输入值以及与神经网络模型中使用的输出节点的耦合系数和偏置来学习语音数据的学习步骤。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号