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

Deep learning characteristic generalization method based on latent variable model

A deep learning and latent variable technology, applied in the field of deep learning feature generalization based on latent variables, which can solve problems such as low image quality

Inactive Publication Date: 2018-09-04
HANGZHOU DIANZI UNIV
View PDF0 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This type of method is relatively simple in modeling, the training process is stable and controllable, and the convergence is fast, but the quality of the generated image is slightly lower than the former

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
  • Deep learning characteristic generalization method based on latent variable model
  • Deep learning characteristic generalization method based on latent variable model
  • Deep learning characteristic generalization method based on latent variable model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0104] The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings.

[0105] A preferred data flow processing method of the present invention is as follows: Image 6 As shown, the specific implementation method is as follows:

[0106] First, the original DNN needs to be divided into two parts, DNN-1 and DNN-2, where X is the feature map output by DNN-1, and its dimension is expressed as:

[0107] x dim =F num ×Size height ×Size width (32)

[0108] f num Indicates the number of current feature maps, Size height 、Size width represent the height and width of a feature map, respectively. As mentioned in the previous section, p(z|x) is a Gaussian form with an approximate diagonal covariance structure, then the posterior probability is expressed as a parameterized Gaussian distribution:

[0109]

[0110] The mapping x→p(z|x) can choose the transformation of the following form to calculate the mean v...

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 discloses a deep learning characteristic generalization method based on a latent variable model. The method is carried out according to the following steps that: S1: on the basis of thelatent variable model, establishing one pair of bidirectional parametric mapping between data space and latent variable space, and combining with a weighted relationship to construct a characteristicgeneralization layer; S2: embedding the characteristic generalization layer into a deep neural network, and dividing the network into three parts, including a DNN (Deep Neural Network)-1, the characteristic generalization layer and a DNN-2; S3: determining the optimization objective of the model, and defining a target function; and S4: lowering characteristic pattern data complexity, and establishing a multi-branch parallel forward propagation structure. The method is favorable for improving the generalization ability of a deep network model, and an overfitting phenomenon during small-scale data training can be lightened. Compared with other methods which use a generative model to enhance data, the method disclosed by the invention simplifies network complexity and improves training efficiency.

Description

technical field [0001] The invention belongs to the technical field of machine learning, and in particular relates to a deep learning feature generalization method based on hidden variables. Background technique [0002] In the face of small-scale data sets, it is necessary to generalize existing data, eliminate model overfitting defects as much as possible, and obtain a high-performance deep neural network classification / regression model. Existing data generalization methods are mainly divided into three categories: data generalization methods based on generative adversarial networks, data generalization methods based on geometric transformations, and data generalization methods based on statistical models. [0003] The Generative Adversarial Network consists of a generator and a discriminator. The former generates a sample from a random vector, and the latter identifies the generated sample and the authenticity of the training set sample. When training the generator, the ...

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): G06N3/04G06K9/62
CPCG06N3/045G06F18/29
Inventor 郭春生李睿哲
Owner HANGZHOU DIANZI UNIV
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