Two-stage audio search method

An audio, stage technology, applied in the field of two-stage audio retrieval, can solve problems such as difficult audio processing

Inactive Publication Date: 2009-02-11
ZHEJIANG UNIV
View PDF0 Cites 38 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although these works have also made some noteworthy progress, in general, the task of processing unannotated audio is still difficult due to the high dimensionality of the audio feature space, the subjectivity and ambiguity of the content similarity depending on the user and the query. difficult

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
  • Two-stage audio search method
  • Two-stage audio search method
  • Two-stage audio search method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0153] 7335 audio data were collected through the Internet, roughly divided into four categories:

[0154] 1) Pure music: 2147 audios of pure music were downloaded from the Internet, and each audio was annotated with the name of the song and the instrument.

[0155] 2) Popular music: 3496 audios of popular music were obtained from the Internet, and each audio was annotated with the name of the song, singer and lyrics.

[0156] 3) Public Speaking: This database contains 234 audios of public speaking, using resources available on the Learning English as a Second Language website.

[0157] 4) TV Shows: This database contains the audio of 1458 TV shows from entertainment sites, each audio is annotated with performer name, show title and some content scripts.

[0158] The parameters used in the embodiment are set as follows:

[0159] 1) For the truncation threshold γ∈(0, 1), choose γ=0.2, 0.4, ..., 1.0, for the weight sequence {κ i |i=1,...}, select candidate sequence ...

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 discloses a two-stage audio search method, which comprises the following steps: 1) extracting the audio features of an audio file in a database; 2) carrying out text-based search for the audio file in the database; 3) composing a training set based on the searched audio file, and searching for a characteristic set which is most reliable for sorting by adopting a principal component analysis method; 4) based on the training set, training the weak classifier composed of the characteristic set so as to compose a strong classifier; and 5) searching by the strong classifier obtained in step 4). The two-stage audio search method is suitable for any content-based recommendation system of an audio set, the searched part of which is remarked, and is also suitable for any non-text objects, such as images and videos.

Description

technical field [0001] The invention relates to the field of computer Internet multimedia search, in particular to a two-stage audio search method. Background technique [0002] Today's information retrieval techniques have achieved great success in processing text documents, as evidenced by the huge commercial profits of search engine companies such as Google (Google) and Yahoo (Yahoo!). In contrast, multimedia retrieval technology is still in its infancy, and there are no products or tools that can achieve the user satisfaction and popularity achieved by text-based search engines. In fact, the problem of retrieving unannotated audio has received less attention than its importance and widespread application. [0003] Existing recommender systems rely heavily on textual annotations [1] when processing audio data. These annotations contain structured or unstructured metadata such as title, artist, and lyrics. The method of retrieving audio based on text annotations is esse...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06F17/30
Inventor 徐颂华陈苏超秦学英刘智满潘云鹤
Owner ZHEJIANG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap