Speech recognition method and electronic device thereof

Abstract

The invention relates to a speech recognition method and an electronic device of the speech recognition method. The speech recognition method includes the steps of inputting a feature vector acquired from a speech signal into a plurality of speech recognition modules, acquiring a plurality of word string probabilities and a plurality of candidate word strings from the speech recognition modules respectively, and selecting the candidate word string corresponding to the largest word string probability as the recognition result of the speech signal. With the speech recognition method, a language corresponding to the speech signal can be automatically recognized.

Description

Technical field [0001] The present invention relates to a voice recognition technology, and more particularly to a voice recognition method and electronic device that can be used to recognize different languages. Background technique [0002] Speech recognition is undoubtedly a hot research and business topic. Speech recognition usually takes out the characteristic parameters of the input speech and compares it with the samples in the database to find out the samples with low degree of dissimilarity from the input. [0003] At present, most common methods are to collect speech corpus (such as the recorded voice of a person) first, and then manually label it (that is, to label the corresponding text on each sentence), and then use these corpus to train acoustic models and acoustic dictionaries. The acoustic model is a statistical classifier. The current practice often uses the Gaussian Mixture Model, which classifies the input speech into basic phonemes. Phonemes are the basic ph...

AI Technical Summary

Problems solved by technology

Question 1: If the user's non-standard pronunciation (such as no distinction between tongue-turned sound, front-back nasal sound, etc.) enters the acoustic model, the ambiguity of the acoustic model will increase

For example, the pinyin "in" will give a relatively high probability of "ing" in the acoustic model, and this compromise for non-standard pronunciation will lead to an increase in the overall error rate

Question 2: Due to the different pronunciation habits in different regions, non-standard pronunciation has many deformations, which leads to greater ambiguity of the acoustic model, thus further reducing the recognition accuracy

Problem 3: Unable to recognize dialects, such as standard Mandarin, Shanghainese, Cantonese, Hokkien, etc.

Images