Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method and device for establishing picture search correlation prediction model, and picture search method and device

A technology of image search and prediction model, applied in the field of information processing, can solve the problems of lack of statistical significance, sparse clicks, and no clicks, etc., to achieve the effect of optimizing image search technology and improving relevance

Active Publication Date: 2016-10-12
BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD
View PDF4 Cites 79 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The above three methods for describing the relevance characteristics of image search all have certain limitations:
[0007] Click-through rate characteristics: mainly based on user behavior statistics. On the one hand, there are biases and noises, and on the other hand, there is sparsity. Only pictures that are displayed at the top and have a sufficient number of times under high-frequency queries can count sufficient clicks. In other cases , no clicks are counted, or the clicks are very sparse, lacking statistical significance

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 and device for establishing picture search correlation prediction model, and picture search method and device
  • Method and device for establishing picture search correlation prediction model, and picture search method and device
  • Method and device for establishing picture search correlation prediction model, and picture search method and device

Examples

Experimental program
Comparison scheme
Effect test

no. 1 example

[0042] figure 1It is a flow chart of a method for establishing a picture search correlation prediction model provided in the first embodiment of the present invention. The method of this embodiment can be executed by the device for establishing a picture search correlation prediction model. The device can use hardware and / or or software, and generally can be integrated into a server used to establish a picture search correlation prediction model. The method of this embodiment specifically includes:

[0043] 110. Using the training samples to train the pre-built raw deep neural network.

[0044] In this embodiment, the training samples include: query formula and picture data.

[0045] As mentioned above, in order to realize that the final output of the deep neural network is the correlation measure between the picture and the query expression, it is necessary to use the image data and the query expression as training samples to train the original deep neural network.

[0046...

no. 2 example

[0092] Figure 4 It is a flowchart of a method for establishing a picture search correlation prediction model according to the second embodiment of the present invention. This embodiment is optimized on the basis of the above-mentioned embodiments. In this embodiment, the specific optimization of using training samples to train the pre-built original deep neural network is as follows: select a set number of training samples; sequentially obtain a training sample input to the original deep neural network, and adjust the weighting parameters in the original deep neural network according to the output of the original deep neural network based on the training sample; return to execute to obtain a training sample input to the The operation of the original deep neural network until the pre-set training end condition is reached.

[0093] Correspondingly, the method in this embodiment specifically includes:

[0094] 410. Select a set number of training samples.

[0095] Considering...

no. 3 example

[0113] Figure 6 It is a flowchart of a method for establishing a picture search correlation prediction model according to the third embodiment of the present invention. This embodiment is optimized on the basis of the above embodiments. In this embodiment, the specific optimization of selecting a set number of training samples is as follows: according to the image click log of the search user, summarize the image click information corresponding to the same query sample, Wherein, the query sample includes: a single query or at least two query that meet the set similarity threshold condition; according to the aggregated click information on the picture, a positive image sample set corresponding to the query sample is generated; and Negative picture sample set; select the query formula sample of setting quantity as described training query formula, and according to the positive picture sample set and the negative picture sample set respectively corresponding to each described tr...

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 embodiment of the invention discloses a method and a device for establishing a picture search correlation prediction model, and a picture search method and device. The method for establishing the picture search correlation prediction model comprises the following steps: using a training sample to train a pre-constructed original deep neural network, wherein the training sample comprises a query and picture data, and the original deep neural network comprises a representation vector generation network and a relevant computational network; and taking the original deep neural network which finishes training as the picture search correlation prediction model. The technical scheme of the invention optimizes the traditional picture search technology, and is better than the traditional technology and various fusion and variation capabilities on multiple aspects including the semantic matching of the query and a picture text, the semantic matching of the query and picture contents, click generalization and the like, and relevancy between a picture search result and the query input by the user can be greatly improved.

Description

technical field [0001] Embodiments of the present invention relate to information processing technologies, and in particular, relate to establishment of a picture search correlation prediction model, a picture search method and device. Background technique [0002] Image search refers to the user's input of natural language queries, for example, through the query formula (also called Query) entered in the text input box provided by the search engine, searching from the image collection and returning the sorted image results according to the relevance and other indicators Information retrieval process for users. Relevance is one of the most important performance indicators of search engines, which measures the degree of relevance between returned results and user query requirements. For an image search engine, the returned image is in an unstructured pixel format, while the query entered by the user is in a text format. These are two completely different information formats,...

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): G06F17/30G06N3/08
CPCG06F16/5866G06N3/08G06F16/56G06F16/9535G06V10/82G06N3/045G06F18/2148G06N3/047
Inventor 付立波罗恒方高林徐伟
Owner BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products