Voice recognition system for mobile terminal and method

Abstract

The invention discloses a voice recognition system for a mobile terminal and a method. The voice recognition system for the mobile terminal comprises the mobile terminal and a server side. The mobile terminal comprises a voice collecting module, a mobile terminal communication module, a user behavior collecting module, a mobile terminal voice recognition module, a result comparison module and a result output module. The server side comprises a server side communication module, a server side voice recognition module, a postprocessing module and a voice module generating module. The mobile terminal recognizes the voice of a user through the mobile terminal voice recognition module, the server side recognizes the voice of the user through the server side voice recognition module, and the result comparison module receives the recognition results of the mobile terminal and the server side and compares the recognition results; the final voice recognition result is sent to the result output module, the result output module carries out corresponding operation according to the final voice recognition result, and convenience is brought to intelligent control over the mobile terminal by the user.

Description

technical field [0001] The present invention relates to the technical field of voice recognition, in particular to a voice recognition system and method for a mobile terminal. Background technique [0002] Speech recognition technology is a high technology that allows machines to convert voice signals into corresponding text or commands through the process of recognition and understanding. Speech recognition (speech recognition) is a technology that converts speech into text and performs corresponding operations. Using speech recognition technology on various smart devices for human-computer interaction and performing various operations has become an important function. [0003] There are two basic methods of speech recognition in the prior art, one is embedded speech recognition (embedded speech recognition), this method is generally recognized by the mobile terminal itself, and the embedded speech recognition directly performs speech recognition on the smart terminal. The...

AI Technical Summary

Problems solved by technology

[0003] There are two basic methods of speech recognition in the prior art, one is embedded speech recognition (embedded speech recognition), this method is generally recognized by the mobile terminal itself, and the embedded speech recognition directly performs speech recognition on the smart terminal. The network is required, and the main calculation is performed on the client side, so it has a fast response time. In the current embedded system, especially some smart devices that use power instead of batteries (such as smart TVs), strong calculations can be achieved Ability to handle some speech recognition tasks, but complex tasks are more difficult to complete

[0004] The second is voice recognition based on cloud-based speech recognition. Generally, after the mobile terminal receives the user's voice, it sends it to the background server for recognition. The cloud voice recognition sends the voice to the powerful background server through the network, and the server performs Return the result after processing. By transmitting the voice or voice acoustic features on the smart device to the server, and using the powerful computing power of the cloud server, more complex recognition tasks can be realized. The disadvantage is that the network and processing delay are relatively large, that is, the feedback Speech recognition results take a long time

[0005] If you want to use embedded-based speech recognition and cloud-based speech recognition at the same time, you need a more complicated coordination method, and there is no good solution yet

Images