Abstract:
The exact data used to train our deep convolutional neural networks (see our [research page](http://www.robots.ox.ac.uk/~vgg/research/text/)) is included in this torrent. This is synthetically generated dataset which we found sufficient for training text recognition on real-world images  This dataset consists of *9 million images* covering *90k English words*, and includes the training, validation and test splits used in our work.
Year: 2016
Journal: International Journal of Computer Vision, Volume: 116, Pages: 1--20

URL: https://www.robots.ox.ac.uk/~vgg/data/text/
License: No license specified, the work may be protected by copyright.

@article{jaderberg16,
author= {Max Jaderberg and Karen Simonyan and Andrea Vedaldi and Andrew Zisserman},
title= {Reading Text in the Wild with Convolutional Neural Networks},
journal= {International Journal of Computer Vision},
number= {1},
volume= {116},
pages= {1--20},
month= {jan},
year= {2016},
abstract= {The exact data used to train our deep convolutional neural networks (see our [research page](http://www.robots.ox.ac.uk/~vgg/research/text/)) is included in this torrent.

This is synthetically generated dataset which we found sufficient for training text recognition on real-world images

![Synthetic Data Engine processt](https://i.imgur.com/cqmgbUa.png)

This dataset consists of *9 million images* covering *90k English words*, and includes the training, validation and test splits used in our work.},
keywords= {},
terms= {},
license= {},
superseded= {},
url= {https://www.robots.ox.ac.uk/~vgg/data/text/}
}