Voice query method and device, computer equipment and storage medium

Abstract

The invention discloses a voice query method and device, computer equipment and a storage medium, and the method comprises the steps: receiving a scene query instruction containing a scene identifier,and obtaining source voice data; inputting the trained deep learning pinyin recognition model, and outputting pinyin sequence data according to the extracted voice pinyin features; searching a keyword pinyin string matched with the pinyin string contained in the pinyin sequence data from a pinyin dictionary matched with the scene identifier, replacing the keyword pinyin string with a keyword corresponding to the keyword pinyin string matched with the keyword pinyin string, and recording the keyword pinyin string as pinyin conversion data; inputting the pinyin conversion data into a language recognition model, predicting predicted Chinese characters according to word meaning characteristics, and obtaining final voice data; and inputting, obtaining a query result through the SQL statement conversion model, and displaying the query result. According to the invention, professional terms in the voice data of the user can be quickly, accurately and automatically recognized, and the recognition accuracy and the user experience satisfaction are improved.

Description

technical field [0001] The invention relates to the field of voice signal processing, in particular to a voice query method, device, computer equipment and storage medium. Background technique [0002] At present, in the field of Internet technology, due to the ever-changing business scenarios and new business scenarios, more and more data are stored in the database, and there are more and more professional terms for different business scenarios. Therefore, the professional terms are recognized through speech recognition. Making accurate queries is increasingly becoming an imperative. In the prior art, the speech recognition system mainly uses a large number of training word samples to train the deep neural network to realize speech recognition. Since the evaluation rate in the training word samples is extremely low or none, the professional The speech recognition system often converts it into homonyms, but it is not the correct professional term. For example, the user said...

AI Technical Summary

Problems solved by technology

In the prior art, the speech recognition system mainly uses a large number of training word samples to train the deep neural network to realize speech recognition. Since the evaluation rate in the training word samples is extremely low or none, the professional The speech recognition system often converts it into homonyms, but it is not the correct professional term. For example, the user said "property consultant", but the speech recognition system converted it into "career consultant", etc., resulting in the result obtained through the voice query. The really desired result leads to low recognition accuracy and low user experience satisfaction

Images