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

Overview

Overview

This collection demonstrates how to construct and train a deep, bidirectional stacked LSTM using CNN features as input with CTC loss to perform robust word recognition.

The model is a straightforward adaptation of Shi et al.'s CRNN architecture (arXiv:1507.0571). The provided code downloads and trains using Jaderberg et al.'s synthetic data (IJCV 2016), MJSynth.

Notably, the model achieves a lower test word error rate (1.82%) than CRNN when trained and tested on case-insensitive, closed vocabulary MJSynth data.

Written for Python 2.7. Requires TensorFlow >=1.10 (deprecation warnings exist for TF>1.10, but the code still works).

The model and subsequent experiments are more fully described in Weinman et al. (ICDAR 2019)

Structure

The model as built is a hybrid of Shi et al.'s CRNN architecture (arXiv:1507.0571) and the VGG deep convnet, which reduces the number of parameters by stacking pairs of small 3x3 kernels. In addition, the pooling is also limited in the horizontal direction to preserve resolution for character recognition. There must be at least one horizontal element per character.

Assuming one starts with a 32x32 image, the dimensions at each level of filtering are as follows:

Layer Op KrnSz Stride(v,h) OutDim H W PadOpt
1 Conv 3 1 64 30 30 valid
2 Conv 3 1 64 30 30 same
Pool 2 2 64 15 15
3 Conv 3 1 128 15 15 same
4 Conv 3 1 128 15 15 same
Pool 2 2,1 128 7 14
5 Conv 3 1 256 7 14 same
6 Conv 3 1 256 7 14 same
Pool 2 2,1 256 3 13
7 Conv 3 1 512 3 13 same
8 Conv 3 1 512 3 13 same
Pool 3 3,1 512 1 13
9 LSTM 512
10 LSTM 512

To accelerate training, a batch normalization layer is included before each pooling layer and ReLU non-linearities are used throughout. Other model details should be easily identifiable in the code.

The default training mechanism uses the ADAM optimizer with learning rate decay.

Differences from CRNN

Deeper early convolutions

The original CRNN uses a single 3x3 convolution in the first two conv/pool stages, while this network uses a paired sequence of 3x3 kernels. This change increases the theoretical receptive field of early stages of the network.

As a tradeoff, we omit the computationally expensive 2x2x512 final convolutional layer of CRNN. In its place, this network vertically max pools over the remaining three rows of features to collapse to a single 512-dimensional feature vector at each horizontal location.

The combination of these changes preserves the theoretical receptive field size of the final CNN layer, but reduces the number of convolution parameters to be learned by 15%.

Padding

Another important difference is the lack of zero-padding in the first convolutional layer, which can cause spurious strong filter responses around the border. By trimming the first convolution to valid regions, this model erodes the outermost pixel of values from the response filter maps (reducing height from 32 to 30 and reducing the width by two pixels).

This approach seems preferable to requiring the network to learn to ignore strong Conv1 responses near the image edge (presumably by weakening the power of filters in subsequent convolutional layers).

Batch normalization

We include batch normalization after each pair of convolutions (i.e., after layers 2, 4, 6, and 8 as numbered above). The CRNN does not include batch normalization after its first two convolutional stages. Our model therefore requires greater computation with an eye toward decreasing the number of training iterations required to reach converegence.

Subsampling/stride

The first two pooling stages of CRNN downsample the feature maps with a stride of two in both spatial dimensions. This model instead preserves sequence length by downsampling horizontally only after the first pooling stage.

Because the output feature map must have at least one timeslice per character predicted, overzealous downsampling can make it impossible to represent/predict sequences of very compact or narrow characters. Reducing the horizontal downsampling allows this model to recognize words in narrow fonts.

This increase in horizontal resolution does mean the LSTMs must capture more information. Hence this model uses 512 hidden units, rather than the 256 used by the CRNN. We found this larger number to be necessary for good performance.

Training

To completely train the model, you will need to download the mjsynth dataset and pack it into sharded TensorFlow records. Then you can start the training process, a tensorboard monitor, and an ongoing evaluation thread. The individual commands are packaged in the accompanying Makefile.

make mjsynth-download
make mjsynth-tfrecord
make train &
make monitor &
make test

To monitor training, point your web browser to the url (e.g., (http://127.0.1.1:8008)) given by the Tensorboard output.

Note that it may take 4-12 hours to download the complete mjsynth data set. A very small set (0.1%) of packaged example data is included; to run the small demo, skip the first two lines involving mjsynth.

With a GeForce GTX 1080, the demo takes about 20 minutes for the validation character error to reach 45% (using the default parameters); at one hour (roughly 7000 iterations), the validation error is just over 20%.

With the full training data, by one million iterations the model typically converges to around 5% training character error and 27.5% word error.

Checkpoints

Pre-trained model checkpoints at DOI:11084/23328 are used to produce results in the following paper:

Weinman, J. et al. (2019) Deep Neural Networks for Text Detection and Recognition in Historical Maps. In Proc. ICDAR.

Testing

The evaluate script (src/evaluate.py) streams statistics for one batch of validation (or evaluation) data. It prints the iteration, evaluation batch loss, label error (percentage of characters predicted incorrectly), and the sequence error (percentage of words—entire sequences—predicted incorrectly).

The test script (src/test.py) tallies statistics, finally normalizing for all data. It prints the loss, label error, total number of labels, sequence error, total number of sequences, and the label error rate and sequence error rate.

Validation

To see the output of a small set of instances, the validation script (src/validation.py) allows you to load a model and read an image one at a time via the process's standard input and print the decoded output for each. For example

cd src ; python validate.py < ~/paths_to_images.txt

Alternatively, you can run the program interactively by typing image paths in the terminal (one per line, type Control-D when you want the model to run the input entered so far).

Configuration

There are many command-line options to configure training parameters. Run train.py or test.py with the --help flag to see them or inspect the scripts. Model parameters are not command-line configurable and need to be edited in the code (see src/model.py).

Dynamic training data

Dynamic data can be used for training or testing by setting the --nostatic_data flag.

You can use the --ipc_synth boolean flag [default=True] to determine whether to use single-threaded or a buffered, multiprocess synthesis.

The --synth_config_file flag must be given with --nostatic_data.

The MapTextSynthesizer library supports training with dynamically synthesized data. The relevant code can be found within MapTextSynthesizer/tensorflow/generator

Using a lexicon

By default, recognition occurs in "open vocabulary" mode. That is, the system observes no constraints on producing the resulting output strings. However, it also has a "closed vocabulary" mode that can efficiently limit output to a given word list as well as a "mixed vocabulary" mode that can produce either a vocabulary word from a given word list (lexicon) or a non-vocabulary word, depending on the value of a prior bias for lexicon words.

Using the closed or mixed vocabulary modes requires additional software. This repository is connected with a fork of Harald Scheidl's CTCWordBeamSearch, obtainable as follows:

git clone https://github.com/weinman/CTCWordBeamSearch
cd CTCWordBeamSearch
git checkout var_seq_len

Then follow the build instructions, which may be as simple as running

cd cpp/proj
./buildTF.sh

To use, make sure CTCWordBeamSearch/cpp/proj (the directory containing TFWordBeamSearch.so) is in the LD_LIBRARY_PATH when running test.py or validate.py (in this repository).

API Notes

This version uses the TensorFlow (v1.14) Dataset for fast I/O. Training, testing, validation, and prediction use a custom Estimator.

Citing this work

Please cite the following paper if you use this code in your own research work:

@inproceedings{ weinman19deep,
    author = {Jerod Weinman and Ziwen Chen and Ben Gafford and Nathan Gifford and Abyaya Lamsal and Liam Niehus-Staab},
    title = {Deep Neural Networks for Text Detection and Recognition in Historical Maps},
    booktitle = {Proc. IAPR International Conference on Document Analysis and Recognition},
    month = {Sep.},
    year = {2019},
    location = {Sydney, Australia},
    doi = {10.1109/ICDAR.2019.00149}
} 

Acknowledgment

This work was supported in part by the National Science Foundation under grant Grant Number 1526350.

Owner
Jerod Weinman
Associate Professor of Computer Science
Jerod Weinman
Python-based tools for document analysis and OCR

ocropy OCRopus is a collection of document analysis programs, not a turn-key OCR system. In order to apply it to your documents, you may need to do so

OCRopus 3.2k Dec 31, 2022
Code for generating synthetic text images as described in "Synthetic Data for Text Localisation in Natural Images", Ankush Gupta, Andrea Vedaldi, Andrew Zisserman, CVPR 2016.

SynthText Code for generating synthetic text images as described in "Synthetic Data for Text Localisation in Natural Images", Ankush Gupta, Andrea Ved

Ankush Gupta 1.8k Dec 28, 2022
Learning Camera Localization via Dense Scene Matching, CVPR2021

This repository contains code of our CVPR 2021 paper - "Learning Camera Localization via Dense Scene Matching" by Shitao Tang, Chengzhou Tang, Rui Hua

tangshitao 65 Dec 01, 2022
MXNet OCR implementation. Including text recognition and detection.

insightocr Text Recognition Accuracy on Chinese dataset by caffe-ocr Network LSTM 4x1 Pooling Gray Test Acc SimpleNet N Y Y 99.37% SE-ResNet34 N Y Y 9

Deep Insight 99 Nov 01, 2022
Recognizing cropped text in natural images.

ASTER: Attentional Scene Text Recognizer with Flexible Rectification ASTER is an accurate scene text recognizer with flexible rectification mechanism.

Baoguang Shi 681 Jan 02, 2023
Deep Learning Chinese Word Segment

引用 本项目模型BiLSTM+CRF参考论文:http://www.aclweb.org/anthology/N16-1030 ,IDCNN+CRF参考论文:https://arxiv.org/abs/1702.02098 构建 安装好bazel代码构建工具,安装好tensorflow(目前本项目需

2.1k Dec 23, 2022
This is the official PyTorch implementation of the paper "TransFG: A Transformer Architecture for Fine-grained Recognition" (Ju He, Jie-Neng Chen, Shuai Liu, Adam Kortylewski, Cheng Yang, Yutong Bai, Changhu Wang, Alan Yuille).

TransFG: A Transformer Architecture for Fine-grained Recognition Official PyTorch code for the paper: TransFG: A Transformer Architecture for Fine-gra

Ju He 307 Jan 03, 2023
~1000 book pages + OpenCV + python = page regions identified as paragraphs, lines, images, captions, etc.

cosc428-structor I had an open-ended Computer Vision assignment to complete, and an out-of-copyright book that I wanted to turn into an ebook. Convent

Chad Oliver 45 Dec 06, 2022
Rest API Written In Python To Classify NSFW Images.

✨ NSFW Classifier API ✨ Rest API Written In Python To Classify NSFW Images. Fastest Solution If you don't want to selfhost it, there's already an inst

Akshay Rajput 23 Dec 30, 2022
Creating a virtual tv using opencv in python3.

Virtual-TV Creating a virtual tv using opencv in python3. In order to run the code follow the below given steps: Make sure the desired videos which ar

Vamsi 1 Jan 01, 2022
Official implementation of Character Region Awareness for Text Detection (CRAFT)

CRAFT: Character-Region Awareness For Text detection Official Pytorch implementation of CRAFT text detector | Paper | Pretrained Model | Supplementary

Clova AI Research 2.5k Jan 03, 2023
Repository for Scene Text Detection with Supervised Pyramid Context Network with tensorflow.

Scene-Text-Detection-with-SPCNET Unofficial repository for [Scene Text Detection with Supervised Pyramid Context Network][https://arxiv.org/abs/1811.0

121 Oct 15, 2021
kaldi-asr/kaldi is the official location of the Kaldi project.

Kaldi Speech Recognition Toolkit To build the toolkit: see ./INSTALL. These instructions are valid for UNIX systems including various flavors of Linux

Kaldi 12.3k Jan 05, 2023
pyntcloud is a Python library for working with 3D point clouds.

pyntcloud is a Python library for working with 3D point clouds.

David de la Iglesia Castro 1.2k Jan 07, 2023
Code for the "Sensing leg movement enhances wearable monitoring of energy expenditure" paper.

EnergyExpenditure Code for the "Sensing leg movement enhances wearable monitoring of energy expenditure" paper. Additional data for replicating this s

Patrick S 42 Oct 26, 2022
Detect the mathematical formula from the given picture and the same formula is extracted and converted into the latex code

Mathematical formulae extractor The goal of this project is to create a learning based system that takes an image of a math formula and returns corres

6 May 22, 2022
Contextual speed detection for python

Speed Prediction using Optical Flow and 2D CNN About the challenge: Comma.AI Speed Challenge This challenge was developed by Comma.AI to predict the s

Mahimana Bhatt 2 Dec 16, 2021
Official code for "Bridging Video-text Retrieval with Multiple Choice Questions", CVPR 2022 (Oral).

Bridging Video-text Retrieval with Multiple Choice Questions, CVPR 2022 (Oral) Paper | Project Page | Pre-trained Model | CLIP-Initialized Pre-trained

Applied Research Center (ARC), Tencent PCG 99 Jan 06, 2023
🔎 Like Chardet. 🚀 Package for encoding & language detection. Charset detection.

Charset Detection, for Everyone 👋 The Real First Universal Charset Detector A library that helps you read text from an unknown charset encoding. Moti

TAHRI Ahmed R. 332 Dec 31, 2022
This is a repository to learn and get more computer vision skills, make robotics projects integrating the computer vision as a perception tool and create a lot of awesome advanced controllers for the robots of the future.

This is a repository to learn and get more computer vision skills, make robotics projects integrating the computer vision as a perception tool and create a lot of awesome advanced controllers for the

Elkin Javier Guerra Galeano 17 Nov 03, 2022