Method and device for multi-accent speech recognition

Abstract

The invention discloses a method and a device for multi-accent speech recognition. The method for multi-accent speech recognition comprises the steps: adding an adaptive layer for learning accent-related feature information in a coding stage for a single speech recognition system, enabling an accent representation vector to serve as guidance information for each encoder block, inputting into the adaptive layer and guiding a conversion function in the adaptive layer, wherein one encoder is provided with a plurality of encoder blocks which are connected in series; inputting accent irrelevant features into the adaptive layer at the same time; and mixing the accent irrelevant features and the accent representation vector to form accent relevant features. According to the embodiment of the invention, the injection position, accent cardinal number and different types of accent cardinal numbers of the adaptive layer are further discussed so that better accent adaptation is realized.

Description

technical field [0001] The invention belongs to the technical field of model training, and in particular relates to a method and device for multi-accent speech recognition. Background technique [0002] In related technologies, the end-to-end (E2E, End-to-End) Automatic Speech Recognition (ASR) model directly optimizes the probability of the output sequence given the input acoustic features, and has made great progress in various speech corpora. progress. One of the most pressing needs of ASR today is to support multiple accents in a single system, which is often referred to in the literature as multi-accent speech recognition. Difficulties in recognizing accented speech such as phonetics, phonetics, and grammar pose serious challenges to current ASR systems. A simple approach is to build a single ASR model from mixed data (accents from non-native speakers and standard data from native speakers). However, such models often suffer from severe performance degradation due to...

AI Technical Summary

Problems solved by technology

[0004] In the process of implementing this application, the inventor found that the existing technical solutions have the following defects: using a multi-expert system, each expert has a large scale and redundant parameters, and cannot quickly adjust the model according to the difficulty of accent discrimination

In addition, each accent must have an expert system to focus on the relevant information of the accent, and the model has a large amount of data

Images