Training method and system for RNN transducer model, and device

Abstract

The invention belongs to the technical field of electronic signal processing, particularly relates to a training method and system for an RNN transducer model, and a device, and aims to solve the problem that the RNN transducer model cannot learn alignment information of speech data well. The training method comprises the following steps of extracting the features of speech training data to obtaina speech feature sequence; forcedly aligning the speech feature sequence through a GMM-HMM (Gaussian Mixture Model-Hidden Markov Model) model to obtain an alignment annotation, and splicing differentframes of speech features; training the RNN transducer model based on the spliced speech feature sequence and text annotation training data to obtain the probability distribution and the negative logarithmic loss value of each word in a preset word list; acquiring an alignment loss value; carrying out weighted averaging on the alignment loss value and the negative logarithmic loss value to obtaina combined loss value, and updating the parameters of the model through a backward propagation algorithm; and iteratively training the model. Through the adoption of the training method and system, the alignment information of the speech data can be learned accurately.

Description

technical field [0001] The invention belongs to the technical field of electronic signal processing, and in particular relates to a training method, system and device for an end-to-end speech transcription model. Background technique [0002] As the entrance of human-computer interaction, speech recognition is an important research direction in the field of artificial intelligence. Traditional speech recognition methods generally use a Gaussian mixture model-hidden Markov model-based hybrid model (GMM-HMM). There are many components in the entire system, which are trained separately, and the performance cannot meet the requirements. With the in-depth application of deep learning technology in speech recognition, end-to-end speech recognition has achieved remarkable results. In particular, the recently proposed end-to-end speech transcription model (RNN Transducer Model) based on the cyclic neural network not only greatly simplifies the steps of the speech recognition system...

AI Technical Summary

Problems solved by technology

[0003] In order to solve the above-mentioned problems in the prior art, that is, in order to solve the problem that the end-to-end speech transcription model cannot learn the alignment information of speech data well, resulting in low speech recognition accuracy, the first aspect of the present invention proposes an end-to-end A training method for an end-to-end speech transcription model, the method comprising:

Images