Loading patent…
AssigneeHuawei Technologies Co., Ltd.
InventorsYonggang Song, Chengdong Liu
Estimating the term…
This invention relates to a method of automatic recognition for limited speech and its correlative system and device. The main technical content is: Establish the relationship between the voice file and the identification information for a limited range of speech, and then insert the identification information into the voice file, or replace the content of the voice file with the identification information. After the transmitter sends the modified voice file, the receiver receives the modified voice and distinguishes the identification information in the voice. It then extracts the corresponding speech in the predetermined relationship using this identification information. The benefits of the embodiment of this invention include: a lower-cost method for automatically recognizing limited speech; high efficiency speech recognition, no need to wait for the voice file to finish to complete speech recognition; and lastly, a high recognition hit-rate or accuracy rate.
A method for automatic recognition of a speech with a limited range, characterized by comprising the following processes:
A system for automatic recognition of a speech with a limited range, comprising a voice transmitter and a voice receiver, wherein the transmitter includes a voice storage module and a voice transmission module, and the voice receiver includes a corresponding relationship storage module, an extraction module and a search module;
A type of voice transmission device, characterized by said voice transmission device including a voice storage module and a voice transmission module;
A type of voice receiver device, wherein said voice receiver device includes a corresponding relationship storage module, an extraction module and a search module;
Dependencies are inferred from the claim text.