Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Transmedia searching method based on content correlation

A correlation and cross-media technology, applied in special data processing applications, instruments, electrical digital data processing, etc., can solve problems such as single feature space

Inactive Publication Date: 2007-08-22
ZHEJIANG UNIV
View PDF0 Cites 73 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in the aspect of multimedia data analysis and retrieval, typical correlation analysis is rarely used, because this statistical analysis method is to analyze the correlation information between two different variable fields, while the traditional single-modal retrieval technology research is a single eigenspace of modality

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
  • Transmedia searching method based on content correlation
  • Transmedia searching method based on content correlation
  • Transmedia searching method based on content correlation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0083] As shown in Figure 2, some examples of the topological structure of training data sets in low-dimensional isomorphic subspaces are given. The specific steps implemented in this example are described in detail below in conjunction with the method of the present invention, as follows:

[0084] (1) Collect image data and audio data of 7 semantics (birds, dogs, cars, wars, tigers, squirrels, monkeys) as training data sets;

[0085] (2) Use the feature extraction program to extract the HSV color histogram, color aggregation vector CCV and Tamura direction features of the image, construct a 500-dimensional visual feature vector for each image, and construct a 70×500-dimensional vision for 7 semantic categories. Feature matrix

[0086] (3) Use the feature extraction program to extract the four Mpeg compression domain features of the audio centroid (Centroid), attenuation cutoff frequency (Rolloff), spectrum flow (Spectral Flux) and root mean square (RMS);

[0087] (4) The duration...

Embodiment 2

[0094] As shown in Figure 4, a search example of "war", that is, semantics, is given. The specific steps implemented in this example will be described in detail below in conjunction with the method of the present invention, as follows:

[0095] (1) The input is a color picture with the semantics of "war" as a query example, and the system finds the vector representation in the isomorphic subspace corresponding to the picture;

[0096] (2) Use the existing data format conversion method to express the subspace vector corresponding to the query example in polar coordinates;

[0097] (3) Use the general distance function to calculate the distance between this query example and other images and audio in the database, and return the first 10 nearest images and the first 10 nearest audio examples;

[0098] (4) In addition, directly use the underlying content features of the query example without subspace mapping to match the content features of other images in the database, that is, use ...

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

This invention discloses a method for media-crossing searches based on content relativity, which applies the typical relativity analysis to analyze the content characters of different mode media data, maps a visual sense character vector of image data and an auditory character vector of audio data in a low dimension isomorphic sub-space simultaneously by a sub-space mapping algorithm, measures the relativities among different mode data based on a general distance function and modifies the topological structure of a multi-mode data set in the sub-space to increase the cross media search efficiency effectively.

Description

Technical field [0001] The invention relates to multimedia retrieval, in particular to a cross-media retrieval method based on content relevance. Background technique [0002] Content-based multimedia retrieval is a research hotspot in the field of computer vision and information retrieval. Retrieval is achieved by matching the similarity of underlying features such as vision, hearing, or geometry. As early as 1976, McGurk had revealed that the human brain's cognition of external information needs to span and synthesize different sensory information to form a holistic understanding. Recent studies in cognitive neuropsychology have further verified that the cognitive process of the human brain presents the characteristics of cross-media, and the information from different senses such as vision and hearing stimulates each other and interacts to produce cognitive results. Therefore, there is an urgent need to study a cross-media retrieval method that supports different modalities, a...

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 潘云鹤庄越挺吴飞张鸿
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
Eureka Blog
Learn More
PatSnap group products