Method and device for detecting speech keywords as well as retrieval method and system thereof

Abstract

The invention discloses a method and a device for detecting speech keywords as well as a corresponding retrieval method and a system thereof. The system for detecting the speech keywords comprises a VQ code library, a state characteristic library, a state sequence generating unit and a matching unit, wherein the VQ code library stores VQ codes based on a preset acoustical model; the state characteristic library stores a VQ state sequence matrix of an audio file, and the VQ state sequence matrix comprises VQ code numbers and corresponding acoustical probability values of various audio frames of the audio file; the state sequence generating unit is based on the VQ codes in the VQ code library and converts an input keyword into a corresponding VQ state sequence; and the matching unit utilizes the converted VQ state sequence to reduce the VQ state sequence matrix in the state characteristic library and utilizes the VQ state sequence to search the speech keyword in the reduced VQ state sequence matrix. The invention technology can be utilized to rapidly search the speech keywords in the audio file and rapidly find the audio file containing the speech keyword in an audio file library.

Description

technical field [0001] The invention relates to the detection of speech keywords, in particular to a method and device for detecting speech keywords and a corresponding retrieval method and system, which can quickly detect speech keywords from audio files. Background technique [0002] In recent years, more and more information terminal devices, such as PCs (Personal Computers), PDAs (Personal Digital Assistants), mobile phones, and remote controls, have voice input functions, allowing users to input commands or input Key words. That is to say, it is necessary for such an information terminal device to be able to recognize the keyword that the user wants to input from the unknown voice string input by the user, or to detect the keyword input by the user from the existing voice file, for example, in an audio retrieval system Applications. Therefore, how to accurately and quickly detect keywords in speech files is an important issue in speech recognition technology. [0003...

AI Technical Summary

Problems solved by technology

However, due to the large amount of calculation for vector calculation, and the threshold for feature merging is difficult to determine, and the setting of this threshold directly affects whether the obtained feature vector sequence with phonological structure features is accurate or not, which causes calculation The accuracy is not high, and the linear matching must be rescaled during the matching, which increases the amount of calculation

Images