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

An Efficient Method for Quality Evaluation of Image Retrieval Results

An image retrieval and quality evaluation technology, applied in still image data retrieval, metadata still image retrieval, vector format still image data, etc. , Ignore the visual content of pictures and other issues to achieve the effect of accurate prediction

Active Publication Date: 2018-01-30
UNIV OF SCI & TECH OF CHINA
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

There is a major problem in these studies: the visual content of the picture is ignored and only the text information is considered, but the text information often contains a lot of "noise"
However, these methods still have deficiencies, they lack to take the user input text query into account, and thus have not done a good study on the potential relationship between the user input query and the returned images.

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
  • An Efficient Method for Quality Evaluation of Image Retrieval Results
  • An Efficient Method for Quality Evaluation of Image Retrieval Results
  • An Efficient Method for Quality Evaluation of Image Retrieval Results

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0047] figure 1 It is a flow chart of an efficient method for evaluating the quality of image retrieval results provided by an embodiment of the present invention. like figure 1 As shown, the method mainly includes the following steps:

[0048] Step 11. Reconstruct a visual query Q' according to the retrieved image list L returned by the input text query Q to characterize the visual theme contained in the returned retrieved image list L.

[0049] In the embodiment of the present invention, the retrieved image list L that the text-based image retrieval system returns can be expressed as: L={I 1 ,I 2 ,...,I N}, where picture I i Indicates the i-th picture in the list L, i=[1,N].

[0050] In the embodiment of the present invention, rebuilding a visual query Q' based on the returned search image list L mainly includes the following three steps:

[0051] 1) Image visual word representation.

[0052] In the embodiment of the present invention, the bag-of-visual-words model i...

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 an efficient image retrieval result quality evaluation method, which comprises the following steps of rebuilding a visual inquiry Q' to represent visual themes contained in a returned retrieval image list L according to the retrieval image list L returned by an input text inquiry Q; inputting the rebuilt visual inquiry Q' to one image retrieval system based on contents, and acquiring a corresponding retrieval image list L'; respectively building language models of the retrieval image lists L and L', calculating a KL distance between the two language models, and adopting the distance value as a quality evaluation result of the retrieval result. Through the method disclosed by the invention, the image retrieval performance can be efficiency and accurately forecasted.

Description

technical field [0001] The invention relates to the technical field of image retrieval, in particular to an efficient method for evaluating the quality of image retrieval results. Background technique [0002] Quality evaluation of image retrieval results is a frontier issue in the field of image retrieval. This technology attempts to evaluate the quality of the results returned by the retrieval system for a certain query without relevant information (that is, the degree of relevance between the returned results and the query is unknown). The quality evaluation of retrieval results can have a positive impact on users and retrieval systems at the same time. From the user's point of view, this technology can serve as a bridge, enabling users to interact with the retrieval system in order to obtain better retrieval results; for the retrieval system, ideally, if a retrieval system can automatically predict a certain A query performance, then it can automatically adjust its par...

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 Patents(China)
IPC IPC(8): G06F17/30
CPCG06F16/56G06F16/58
Inventor 田新梅贾强槐
Owner UNIV OF SCI & TECH OF CHINA