Supercharge Your Innovation With Domain-Expert AI Agents!

Neural network method for training and calculating similarity hash codes

A neural network and similarity technology, applied in biological neural network models, calculations, neural architectures, etc., can solve problems such as poor results, and achieve a wide range of applications and fast results

Pending Publication Date: 2021-03-26
央视国际网络无锡有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0009] Therefore, the technical problem to be solved in the present invention is to overcome the defects in the prior art that are not effective in practical situations, thereby providing a neural network method for training and calculating similarity hash codes

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
  • Neural network method for training and calculating similarity hash codes
  • Neural network method for training and calculating similarity hash codes
  • Neural network method for training and calculating similarity hash codes

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0032] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

[0033] A neural network method for training and computing similarity hash codes, comprising the following steps:

[0034] S1: if figure 1 As shown, a neural network is used to map a vector z in a high-dimensional space to a vector x in a low-dimensional space.

[0035] The latitude of z is far greater than the latitude of x. Specifically, in one embodiment, if z represents a grayscale image of 256*256 in size, then the latitude of z is 65536 dimensions, and the latitude of x is generally 32 or 64 dimensions .

[0036] The specific form of the neural network can be a multi-layer perceptron, CNN (...

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 the technical field of computer processing, in particular to a neural network method for training and calculating similarity hash codes, and aims to solve the problem of pooreffect in actual conditions in the prior art. The method is characterized by comprising the following steps: S1, mapping a vector z of a high-dimensional space into a vector x of a low-dimensional space by using a neural network; S2, determining a parameter theta of the neural network; and S3, calculating a similarity hash code by using a neural network f (* theta), namely only taking a 0 / 1 binaryvalue as a forced output value. According to the neural network method for training and calculating the similarity hash codes, the problem that the effect is poor when only linear transformation is adopted is solved, the application range is wider, the neural network method is not limited to images, does not depend on manually designed SIFT or SURF features when the images are processed, the speed is higher, and GPU can be used for further acceleration.

Description

technical field [0001] The invention relates to the technical field of computer processing, in particular to a neural network method for training and calculating similarity hash codes. Background technique [0002] In 2010, the paper "The Video Genome" published by Israeli research scholars first proposed the concept of video gene. The paper proposes that for a picture (or image frame), its SIFT feature is first extracted, and the SIFT feature is a 128-dimensional vector. By extracting SIFT features from a large number of images, a large number of SIFT feature point data can be obtained, and then use these SIFT feature data to find 2048 cluster centers through a clustering algorithm, and use the cluster centers as quantization center points. [0003] Then for any image, the SIFT feature vector can be quantized to one of the cluster center points. In the paper, it is proposed to divide the image into four areas: upper left, upper right, lower left, and lower right. In each ...

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): G06K9/62G06N3/04
CPCG06N3/048G06F18/2321G06F18/22G06F18/214
Inventor 张勇朱立松
Owner 央视国际网络无锡有限公司
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com