senlinuc/caffe_ocr alternatives

same CNN+BLSTM+CTC architecture in TensorFlow, choose if you need a TensorFlow implementation instead of Caffe.

CNN+LSTM+CTC based OCR implemented using tensorflow.

Torch7 implementation of RNN+CTC for line OCR, choose if you prefer Torch over Caffe.

using rnn (lstm or gru) and ctc to convert line image into text, based on torch7 and warp-ctc

TensorFlow CNN+LSTM+CTC with GPL license, choose for a maintained TensorFlow baseline.

Tensorflow-based CNN+LSTM trained with CTC-loss for OCR

Adds spatial transformer to CNN+BLSTM+CTC, choose if you need robustness to deformations in TensorFlow.

use STN+CNN+BLSTM+CTC to do OCR

PyTorch reimplementation of multiple scene text recognition models, choose as a research toolbox.

text_recognition_toolbox: The reimplementation of a series of classical scene text recognition papers with Pytorch in a uniform way.

TensorFlow 2 CRNN implementation, use if you need a modern TF2 codebase.

Convolutional Recurrent Neural Network(CRNN) for End-to-End Text Recognition - TensorFlow 2

CNN-only Chinese text recognition without sequence modeling, choose for simpler character-based recognition tasks.

第一届西安交通大学人工智能实践大赛（2018AI实践大赛--图片文字识别）第一名；仅采用densenet识别图中文字

Adds CTPN text detection before DenseNet+CTC recognition, choose for full end-to-end Chinese OCR pipeline.

CTPN + DenseNet + CTC based end-to-end Chinese OCR implemented using tensorflow and keras

Traditional OCR engine with LSTM, choose for a mature standalone tool with broad language support and no training effort.

Tesseract Open Source OCR Engine (main repository)

Computer vision tools for fairseq containing text recognition, choose if you work in the fairseq ecosystem.

Computer vision tools for fairseq, containing PyTorch implementation of text recognition and object detection

Standalone C++ LSTM library focused on OCR, choose for a lightweight embeddable LSTM without Caffe.

A small C++ implementation of LSTM networks, focused on OCR.

PyTorch research project covering text detection and recognition, choose if you need a complete pipeline beyond recognition.

A research project for text detection and recognition using PyTorch 1.2.

Keras/TF CTPN+DenseNet Chinese OCR with detection, alternative end-to-end pipeline.

Chinese text detection and recognition based on CTPN + DENSENET using Keras and Tensor Flow，使用keras和tensorflow基于CTPN+Densenet实现的中文文本检测和识别

Narrower receipt OCR pipeline for SROIE challenge, choose only for receipt data and competition-style tasks.

ICDAR 2019 Robust Reading Challenge on Scanned Receipts OCR and Information Extraction

CNN+BLSTM+CTC specifically for captcha recognition with deployment focus, narrower than general OCR.

[验证码识别-部署] This project is based on CNN+BLSTM+CTC to realize verificationtion. This projeccode identificat is only for deployment models.

CRNN+CTC applied to Devanagari handwritten words, choose for that specific script/handwriting scenario.

Use Convolutional Recurrent Neural Network to recognize the Handwritten Word text image without pre segmentation into words or characters. Use CTC loss Function to train.

Ready-to-use Python OCR engine with many languages, choose for end-user use rather than integrating into a C++ training pipeline.

Ready-to-use OCR with 80+ supported languages and all popular writing scripts including Latin, Chinese, Arabic, Devanagari, Cyrillic and etc.