/WAP-1

Tensorflow implementation for WAP: Watch, attend and parse: An end-to-end neural network based approach to handwritten mathematical expression recognition

Primary LanguagePython

WAP

Tensorflow implementation for WAP: Watch, attend and parse: An end-to-end neural network based approach to handwritten mathematical expression recognition

Introduction

A end-to-end nonlinear OCR or image2txt model based on feature map attention mechanism and GRU/lstm

Installation

  • python2.7
  • tensorflow 1.3

Usage

  • Train

python2 WAP_Model.py