Scene image labeling method based on coarse-fine granularity multi-image multi-label learning

A scene image and multi-label technology, which is applied in the direction of graphic image conversion, image analysis, image data processing, etc., can solve problems that cannot be directly applied, and there is no MIML learning model feature example vector

Pending Publication Date: 2020-07-28
NORTHEASTERN UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But the existing methods used in MIML cannot be directly applied to the MGML learning environment, because the graph does not have the feature example vectors required by the MIML learning model

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
  • Scene image labeling method based on coarse-fine granularity multi-image multi-label learning
  • Scene image labeling method based on coarse-fine granularity multi-image multi-label learning
  • Scene image labeling method based on coarse-fine granularity multi-image multi-label learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0050] The specific implementation manners of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0051] In this embodiment, the scene image labeling method based on coarse-fine-grained multi-image multi-label learning, such as figure 1 shown, including the following steps:

[0052] Step 1: Obtain the original scene image dataset and the corresponding label set;

[0053] In the embodiment of the present invention, the real image data set: PASCAL VISUAL Object Challenge 2012 data set (VOC12) is used as the original scene graph data set; the data set has a total of 1073 images, and each image has corresponding multiple objects, each Objects have a label. The dataset includes a total of 20 types of objects, such as "car", "boat", "dog", "human", "sheep" and "chair".

...

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 provides a scene image labeling method based on coarse-fine granularity multi-image multi-label learning, and relates to the technical field of image classification. The method comprisesthe following steps: firstly, obtaining an original scene image data set and a corresponding label set, and carrying out data preprocessing on the original scene image data set to obtain a multi-image data structure, namely an image packet; defining a graph-kernel-based graph-level score function and a packet-level score function of each label; constructing an objective function based on sortingloss; and optimizing a target function based on sorting loss through a sub-gradient descent algorithm to obtain an optimal weight value of each label, and then constructing a graph-level classifier and a packet-level classifier to realize prediction of a label set of an unknown multi-graph data packet and a label set of graphs in the packet and complete labeling of a scene image. Label predictionis allowed to be carried out on coarse granularity (packet level) and fine granularity (graphs in packets) at the same time based on defined graph level and packet level score functions, and the category of traditional multi-graph multi-label classification is expanded.

Description

technical field [0001] The invention relates to the technical field of image classification, in particular to a scene image labeling method based on coarse-fine grained multi-map multi-label learning. Background technique [0002] With the wide application of photographic equipment and the development of Internet services, a large number of scene images are generated every day. A single scene image in these scene images generally involves multiple topics, and the semantics are relatively complex. At the same time, a large number of scene images generated by the Internet generally do not have classification labels that can fully describe the image content. For these massive scene images with complex semantics and no classification labels, how to use these images to provide relevant services for Internet users is the core task of scene image annotation. [0003] Current scene image annotation techniques always assume that each sample can be represented in the form of one or ...

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
IPC IPC(8): G06K9/62G06K9/00G06T3/40G06T7/11
CPCG06T7/11G06T3/4053G06V20/00G06F18/2431G06F18/214
Inventor 赵宇海王业江印莹
Owner NORTHEASTERN 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