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

Image retrieval method based on multi-scale NetVLAD and deep Hash

An image retrieval, multi-scale technology, applied in the field of target retrieval and computer vision, can solve the problems of reduced retrieval speed, increased dictionary size, etc., to achieve high retrieval speed, reduced feature dimension and complexity, and reduced computational complexity.

Active Publication Date: 2019-09-10
西安华企众信科技发展有限公司
View PDF2 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, it should be pointed out that with the explosive growth of data today, in order not to reduce the retrieval accuracy, the size of the dictionary used has also increased sharply, resulting in a slowdown in retrieval speed

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
  • Image retrieval method based on multi-scale NetVLAD and deep Hash
  • Image retrieval method based on multi-scale NetVLAD and deep Hash
  • Image retrieval method based on multi-scale NetVLAD and deep Hash

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0058] The present invention will be further described below in conjunction with accompanying drawing.

[0059] Image retrieval methods based on multi-scale NetVLAD and deep hashing, such as figure 1 As shown, step 1, training process: input the training samples into the multi-scale convolutional neural network, and obtain the P-layer convolutional feature group Then it undergoes feature fusion to obtain the fused feature X l , after passing through the NetVLAD layer, the pooled feature V is obtained l , and then hash coded to output the final image feature representation Finally, the backpropagation algorithm is used to derive the loss function and optimize all the learnable parameters that appear in the network. The testing process is to input new sample data into the trained network structure to test the network retrieval accuracy.

[0060] Specific steps are as follows:

[0061] Step 1. Obtain the training sample label: the training sample is divided into a query se...

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 relates to an image retrieval method based on multi-scale NetVLAD and depth Hash. According to the method, a multi-scale convolutional neural network based local aggregation descriptor vector method is adopted to carry out feature extraction on pictures in image retrieval. The original output features which only adopt the last convolutional layer are optimized into the features whichare fused by adopting the output features of the plurality of convolutional layers. The fused features not only comprise high-layer semantic features, but also comprise low-layer picture detail information. A Hash layer is added behind the NetVLAD for feature coding, the features become simpler, the feature dimension and complexity are reduced through a Hash coding layer, and the subsequent storage overhead and calculation complexity are remarkably reduced. The speed of image retrieval is increased, and similar images can be quickly and accurately retrieved in a large-scale data set.

Description

technical field [0001] The invention belongs to the fields of computer vision and target retrieval, and relates to an image retrieval method based on multi-scale NetVLAD and deep hashing. Background technique [0002] Image retrieval technology is to retrieve pictures that meet the conditions from the picture database. It has a wide range of application scenarios in real life, such as remote sensing images, security monitoring, search engines, e-commerce, biomedicine, etc., all play a vital role. role. [0003] The commonly used image retrieval methods mainly include text-based and content-based retrieval. The text-based retrieval method uses artificial or semi-supervised learning methods to assign a set of free text to the image to describe the content of the image, and converts the image retrieval into text retrieval through a text retrieval system. Because images contain rich information, text labels often cannot fully represent the image information, and may even fail ...

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): G06F16/583G06N3/04
CPCG06F16/583G06N3/045
Inventor 叶凌智翁立王建中
Owner 西安华企众信科技发展有限公司
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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com