https://github.com/wondervictor/srn

Sequence Recognition Network Implementation by PyTorch
https://github.com/wondervictor/srn

computer-vision deep-learning seq2seq text-recognition

Last synced: over 1 year ago
JSON representation

Sequence Recognition Network Implementation by PyTorch

• Host: GitHub
• URL: https://github.com/wondervictor/srn
• Owner: wondervictor
• Created: 2017-08-28T13:04:49.000Z (almost 9 years ago)
• Default Branch: master
• Last Pushed: 2017-11-21T14:04:01.000Z (over 8 years ago)
• Last Synced: 2025-01-17T08:35:58.925Z (over 1 year ago)
• Topics: computer-vision, deep-learning, seq2seq, text-recognition
• Language: Jupyter Notebook
• Size: 153 KB
• Stars: 3
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Sequence Recognition Network

****

**PyTorch Implementation of SRN (Sequence Recognition Network)**

### Introduction

SRN combines the CNN and Seq2Seq together to optimize the performance of text recognition in natural scene. 

![](images/SRN.png)

for more about the Sequence Recoginition Network, paper is here => [Robust Scene Text Recognition with Automatic Rectification](http://www.cv-foundation.org/openaccess/content_cvpr_2016/papers/Shi_Robust_Scene_Text_CVPR_2016_paper.pdf)

### Developing

**I am still proceeding in the project.** Any bug, please issue

### ChangeLog

**2017-09-11**: Support Mini-Batch

### TODO

1. support beam search