Unlock instant, AI-driven research and patent intelligence for your innovation.

Method for updating classifier parameters for identifying audio content

A content recognition and parameter update technology, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of inability to update classifier parameters, insufficient, and inability to achieve optimal classification.

Active Publication Date: 2009-09-30
SPREADTRUM COMM (SHANGHAI) CO LTD
View PDF6 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In fact, in many classifier applications, the training samples are very limited or insufficient, and the classifier parameters cannot be updated according to the actual test samples, so that the purpose of optimal classification cannot be achieved.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method for updating classifier parameters for identifying audio content
  • Method for updating classifier parameters for identifying audio content
  • Method for updating classifier parameters for identifying audio content

Examples

Experimental program
Comparison scheme
Effect test

Embodiment approach

[0046] where the parameter α i , β i , i=1, 2, 3 determine the strength of the update, and its specific value can be determined arbitrarily, as long as α i +Kβ i =1, i=1,2,3. One implementation is α i = N N + K , β i = 1 N + K , i = 1,2,3 , Among them, N is the size of the original data set, and K is the number of data in the first data set.

[0047] For the data of data set 2, train its own Gaussian mixture model parameters and update the overall Gaussian mixture model parameters, adopt the following method:

[0048] The first step: according to Calculate the Gaussian mixture model parameters generated by these data (add h Gaussian mixture):

[0...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a method for updating classifier parameters for identifying audio content, which comprises the following steps: acquiring new training data; selecting data to obtain a data set one and a data set two; updating parameters of a Gaussian mixture model by utilizing the data set one; and judging whether data quantity of the data set two is greater than a threshold value or not, and updating the whole parameters of the Gaussian mixture model by utilizing the data set two if the data quantity of the data set two is greater than the threshold value. Aiming at the current Gaussian mixture model, the method can update the classifier parameters according to actually tested samples so as to achieve the aim of optimized classification.

Description

【Technical field】 [0001] The invention relates to a classifier parameter update method for audio content recognition, in particular to a parameter update method suitable for a classifier based on a Gaussian mixture model. 【Background technique】 [0002] Audio is an important medium in multimedia, and audio information retrieval technology is an important part of multimedia information retrieval technology. The corresponding existing technologies can refer to Chinese patents 1391211, 1223739 and 1270361 and US patents 5,613,037, 6,292,776 and 5,440,662, etc. . In audio retrieval applications, it is necessary to classify audio data. Its purpose is to distinguish which category the input audio signal belongs to. Common audio categories include voice, background noise, pop music, classical music, etc., and the application of audio content classification It is also very extensive, especially in the field of audio retrieval, audio content classification plays a decisive role, and...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G10L15/14
Inventor 黄鹤云林福辉
Owner SPREADTRUM COMM (SHANGHAI) CO LTD