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

A style character generation method based on a small number of samples

A character generation and style technology, applied in neural learning methods, character and pattern recognition, digital ink recognition, etc., can solve problems such as scribbled input character style, model inability to handle, multiple training samples, etc.

Active Publication Date: 2019-01-08
XIAN JIAOTONG LIVERPOOL UNIV
View PDF6 Cites 25 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the problem of the model is still that it cannot handle the problem of very sloppy input character style, and it needs more training samples

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
  • A style character generation method based on a small number of samples
  • A style character generation method based on a small number of samples
  • A style character generation method based on a small number of samples

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0070] Embodiment: A character generation method based on a small number of sample styles, characterized in that: several (more than 50 types) of handwritten characters are used as handwritten style transfer targets, or several common type (more than 50 types) of characters are used as typefaces The migration target, and a standard font character as the style transfer source, use the image translation model based on the depth generation confrontation network to train a character generation model for character style transfer;

[0071] The character generation model is composed of Content Prototype Encoder, Enc p , Style Reference Encoder, Enc r It is composed of Decoder and Dec, because the two encoders are structured in parallel, so the total number of layers of the network is 12;

[0072] The content prototype encoder Content Prototype Encoder, Enc p The input data is a character with a standard style ( x 0 j ), represents a grayscale picture with a growth width of 64*64 and a valu...

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 style character generation method based on a small number of samples. A style reference character data set composed of several style characters (handwriting style or print style) and a character of a standard font are used as a character content prototype data source, and an image translation model based on a deep generation antagonism network is used to train a charactergeneration model of character style transfer. The model can generate arbitrary characters with the same writing / printing style by using a given few (or even one) characters with a certain style (writing / printing) as the style reference template. The content of the generated character is determined by the typed content prototype (standard font).

Description

Technical field [0001] The invention relates to a method for generating style characters, in particular to a method for generating handwritten or printed characters based on a few sample styles. Background technique [0002] According to the model described in the paper "Automatic generation of large-scale handwriting fonts via style learning" published by ZHLian et al. in SIGGRAPH ASIA 2016, users can input some of their handwritten characters according to the interface provided to obtain some styles of model output Similar characters. For example, the user provides 266 characters to input into the model to get 27533 characters with similar styles. However, the training of this model needs to provide a large amount of labeled data, and requires more standardized input from the user to get the expected results, and the effect is not good under the condition of less labeled data. The focus of the model is to separate the individual strokes of the font, and then input it into 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): G06F17/22G06K9/00G06N3/04G06N3/08
CPCG06N3/08G06F40/126G06V30/333G06V30/36G06V30/287G06N3/045Y02T10/40
Inventor 黄开竹江浩川杨关禹王晨晖张锐
Owner XIAN JIAOTONG LIVERPOOL 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