Multi-dialect speech recognition method, device and apparatus, and storage medium

Abstract

The embodiment of the invention discloses a multi-dialect speech recognition method, device and apparatus, and a storage medium. The method comprises the steps: obtaining an input speech acoustic feature through a multi-dialect encoder, and outputting a first vector code with a fixed length; identifying the first vector code through a dialect identifier to obtain a corresponding dialect vector; and decoding the first vector code through a multi-dialect decoder according to the dialect vector to obtain a recognition text corresponding to the speech acoustic feature. According to the scheme, dialect recognition efficiency is improved, a large amount of sample data is not needed, and recognition accuracy is better than that of an existing scheme.

Description

technical field

[0001] The embodiments of the present application relate to the computer field, and in particular to a multi-dialect speech recognition method, device, device, and storage medium. Background technique

[0002] Dialects usually refer to variants of the same language, associated with a particular geographic region or social group. For example, Arabic has several variants, including Egyptian Arabic, Gulf Arabic, and Modern Standard Arabic, among others. Although there is a certain degree of similarity between the various dialects, there are usually large differences at the language level. The result is that automatic speech recognition systems trained for one dialect perform poorly on another.

[0003] For automatic speech recognition of dialects, separate model training can be performed for each dialect if sufficient sample data is available for each dialect. However, in the case of scarce dialect resources, this method is less feasible. In the prior art, t...

Images