%0 Journal Article %1 jaderberg2016reading %A Jaderberg, Max %A Simonyan, Karen %A Vedaldi, Andrea %A Zisserman, Andrew %D 2016 %J International Journal of Computer Vision %K cnn reading text to:read %N 1 %P 1--20 %R 10.1007/s11263-015-0823-z %T Reading Text in the Wild with Convolutional Neural Networks %U http://dx.doi.org/10.1007/s11263-015-0823-z %V 116 %X In this work we present an end-to-end system for text spotting---localising and recognising text in natural scene images---and text based image retrieval. This system is based on a region proposal mechanism for detection and deep convolutional neural networks for recognition. Our pipeline uses a novel combination of complementary proposal generation techniques to ensure high recall, and a fast subsequent filtering stage for improving precision. For the recognition and ranking of proposals, we train very large convolutional neural networks to perform word recognition on the whole proposal region at the same time, departing from the character classifier based systems of the past. These networks are trained solely on data produced by a synthetic text generation engine, requiring no human labelled data. Analysing the stages of our pipeline, we show state-of-the-art performance throughout. We perform rigorous experiments across a number of standard end-to-end text spotting benchmarks and text-based image retrieval datasets, showing a large improvement over all previous methods. Finally, we demonstrate a real-world application of our text spotting system to allow thousands of hours of news footage to be instantly searchable via a text query.

@article{jaderberg2016reading, abstract = {In this work we present an end-to-end system for text spotting---localising and recognising text in natural scene images---and text based image retrieval. This system is based on a region proposal mechanism for detection and deep convolutional neural networks for recognition. Our pipeline uses a novel combination of complementary proposal generation techniques to ensure high recall, and a fast subsequent filtering stage for improving precision. For the recognition and ranking of proposals, we train very large convolutional neural networks to perform word recognition on the whole proposal region at the same time, departing from the character classifier based systems of the past. These networks are trained solely on data produced by a synthetic text generation engine, requiring no human labelled data. Analysing the stages of our pipeline, we show state-of-the-art performance throughout. We perform rigorous experiments across a number of standard end-to-end text spotting benchmarks and text-based image retrieval datasets, showing a large improvement over all previous methods. Finally, we demonstrate a real-world application of our text spotting system to allow thousands of hours of news footage to be instantly searchable via a text query.}, added-at = {2017-01-20T00:07:23.000+0100}, author = {Jaderberg, Max and Simonyan, Karen and Vedaldi, Andrea and Zisserman, Andrew}, biburl = {https://www.bibsonomy.org/bibtex/22a5c5f9c4d67f42fcc2792dd02dc3e2c/nosebrain}, description = {Reading Text in the Wild with Convolutional Neural Networks | SpringerLink}, doi = {10.1007/s11263-015-0823-z}, interhash = {dda78e08244aa704c659b6a9a38e136d}, intrahash = {2a5c5f9c4d67f42fcc2792dd02dc3e2c}, issn = {1573-1405}, journal = {International Journal of Computer Vision}, keywords = {cnn reading text to:read}, number = 1, pages = {1--20}, timestamp = {2017-01-20T00:07:23.000+0100}, title = {Reading Text in the Wild with Convolutional Neural Networks}, url = {http://dx.doi.org/10.1007/s11263-015-0823-z}, volume = 116, year = 2016 }