Image element matching method and device, model training method and device and data processing method

A technology of image elements and data processing, applied in computing models, neural learning methods, character and pattern recognition, etc., can solve the problems of large-scale image training difficulties, high designer costs, poor flexibility, etc., to solve training difficulties and improve Matching effect, the effect of reducing labor costs

CN113255709APending Publication Date: 2021-08-13ALIBABA GRP HLDG LTD
0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Publication Date
2021-08-13

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses an image element matching method and device, a model training method and device and a data processing method. The problem of searching the image information matched with the given image element is converted into the problem of predicting whether an edge exists between the node corresponding to the given image element and the other nodes in the matching knowledge graph from the matching knowledge graph, so that the matching of the image elements is flexibly realized, the matching effect is improved, and the labor cost is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] This application involves but is not limited to industrial intelligence technology, especially a method for matching image elements, a device and a model training method, a device and a data processing method. Background technique

[0002] In the related art, the matching problem of image elements is solved through templates or rules involved in advance. This method has poor flexibility, poor diversity, and high designer costs.

[0003] In addition, the problem of matching image elements will also be solved by pre-designing graph autoencoders. However, this method is difficult for large-scale graph training, and the visual features of graph nodes are pre-calculated. The matching results are not ideal. Contents of the invention

[0004] The present application provides an image element collocation method, device and model training method, device and data processing method, which can be flexibly adapted to large-scale image training and improve the c...

Examples

Embodiment Construction

[0088] In order to make the purpose, technical solution and advantages of the application clearer, the embodiments of the application will be described in detail below in conjunction with the accompanying drawings. It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined arbitrarily with each other.

[0089] In a typical configuration of the present application, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0090] Memory may include non-permanent storage in computer-readable media, in the form of random access memory (RAM) and / or nonvolatile memory, such as read-only memory (ROM) or flash memory (flashRAM). Memory is an example of computer readable media.

[0091] Computer-readable media, including both permanent and non-permanent, removable and non-removable media, can be implemented by any method or technology for...