Unlock instant, AI-driven research and patent intelligence for your innovation.

A text recognition method and system based on contrastive learning

A text recognition and text technology, applied in the field of text recognition, can solve problems such as poor performance of text recognition methods

Active Publication Date: 2022-07-15
XIAMEN MEIYA PICO INFORMATION
View PDF3 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, current contrastive learning methods are only suitable for image classification, image segmentation, and image recognition that use the entire image as a single instance, and do not perform well in text recognition methods that contain sequential

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 text recognition method and system based on contrastive learning
  • A text recognition method and system based on contrastive learning
  • A text recognition method and system based on contrastive learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] The present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the related invention, but not to limit the invention. In addition, it should be noted that, for the convenience of description, only the parts related to the related invention are shown in the drawings.

[0037] It should be noted that the embodiments in the present application and the features of the embodiments may be combined with each other in the case of no conflict. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0038] figure 1 An exemplary system architecture 100 of a text recognition method based on contrastive learning to which embodiments of the present application can be applied is shown.

[0039] like figure 1As shown, the system architec...

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 text recognition method and system based on contrastive learning, including unlabeled text image samples, each of which is subjected to data enhancement and input to a convolutional network for recognition training to generate a recognition model, and then based on the recognition model Build a basic encoder to calculate and output the feature sequence; input the feature sequence into the instance mapping function to generate corresponding instances and then map them into multiple sub-instances, use all sub-instances as sub-elements in the contrast loss function for comparative learning, and compare the results. Feedback to the convolutional network for updating the convolutional network; obtain a labeled text image sample containing text information and input it into the basic encoder, and adjust the parameters of the convolutional network until the recognition model converges . This method applies comparative learning to each element of the sequence, makes full use of unlabeled data to learn effective representation information, and then conducts modeling based on the method of self-supervised comparative learning, which significantly improves the recognition effect.

Description

technical field [0001] The invention relates to the technical field of text recognition, in particular to a text recognition method and system based on contrastive learning. Background technique [0002] In recent years, deep learning has been widely used in image classification, target detection, segmentation and other fields, and it has also brought problems such as high data labeling cost and long time consumption. How to make good use of a large amount of unlabeled data for self-supervised learning has become a current research hotspots. Self-supervised learning can mine its own supervision information from large-scale unlabeled data sets, and train the network through this constructed supervision information, so that it can learn valuable representation information for downstream tasks. [0003] Techniques that utilize self-supervised contrastive learning to represent information have achieved remarkable results in semi-supervised computer vision applications such as i...

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 Patents(China)
IPC IPC(8): G06V10/22G06V30/10G06V10/74G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/088G06N3/045G06F18/22
Inventor 刘彩玲吴婷婷赵建强高志鹏汪泰伸陈德意
Owner XIAMEN MEIYA PICO INFORMATION