Streaming phonetic transcription system based on self-attention mechanism

Abstract

The invention discloses a streaming phonetic transcription system based on a self-attention mechanism. The streaming phonetic transcription system based on the self-attention mechanism comprises a feature front-end processing module, a self-attention audio coding network module, a self-attention prediction network module and a united network module. The feature front-end processing module is usedfor receiving an input acoustic feature and converting into a vector with specific dimensionality; the self-attention audio coding network module is connected with the feature front-end processing module and is used for receiving the processed acoustic feature and obtaining an coded acoustic state vector; the self-attention prediction network module is used for generating a language state vector according to an input prediction mark of the last moment; and the united network module is connected with the self-attention audio coding network module and the self-attention prediction network module, and is used for combining with an acoustic state and a language state and calculating the probability of a new prediction mark. The invention provides a streaming feedforward voice coder based on the self-attention mechanism, so that the calculation efficiency and the precision of a traditional voice coder are improved.

Description

technical field [0001] The invention relates to the technical field of signal processing in the electronics industry, in particular to a streaming speech transcription system based on a self-attention mechanism. Background technique [0002] Speech is one of the main means for human to exchange information. Speech recognition technology mainly enables computers to recognize human speech and transcribe it into corresponding text. In the development of speech recognition technology, the early research mainly used the method based on the mixed Gaussian model and hidden Markov model. The development of deep neural network made the Gaussian model replaced by deep neural network. In recent years, with the development of computer technology, the end-to-end model has attracted more and more attention because of its simplified process and elegant model structure. [0003] The recurrent neural network speech transcription system uses a recurrent neural network as the basic network fr...

AI Technical Summary

Problems solved by technology

However, the recursive calculation of the cyclic neural network is relatively inefficient during the training process.

The self-attention mechanism can also model long-distance dependencies, but requires a complete sequence as input. Although it has high computational efficiency, it cannot be applied to streaming sequence modeling tasks.

Images