Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Synthetic Data and Artificial Neural Networks for Natural Scene Text Recognition

About

In this work we present a framework for the recognition of natural scene text. Our framework does not require any human-labelled data, and performs word recognition on the whole image holistically, departing from the character based recognition systems of the past. The deep neural network models at the centre of this framework are trained solely on data produced by a synthetic text generation engine -- synthetic data that is highly realistic and sufficient to replace real data, giving us infinite amounts of training data. This excess of data exposes new possibilities for word recognition models, and here we consider three models, each one "reading" words in a different way: via 90k-way dictionary encoding, character sequence encoding, and bag-of-N-grams encoding. In the scenarios of language based and completely unconstrained text recognition we greatly improve upon state-of-the-art performance on standard datasets, using our fast, simple machinery and requiring zero data-acquisition costs.

Max Jaderberg, Karen Simonyan, Andrea Vedaldi, Andrew Zisserman• 2014

Related benchmarks

TaskDatasetResultRank
Scene Text RecognitionSVT (test)
Word Accuracy95.4
289
Scene Text RecognitionIIIT5K (test)
Word Accuracy97.1
244
Scene Text RecognitionSVT 647 (test)
Accuracy68
101
Text DetectionICDAR 2013 (test)
F1 Score84.1
88
Scene Text RecognitionIC03 (test)
Accuracy98.7
63
Scene Text RecognitionIC13 1015 (test)
Accuracy79.5
27
Scene Text RecognitionICDAR case-insensitive 2013 (test)
Accuracy90.8
22
Scene Text RecognitionICDAR 50-word lexicon 2003 (test)
Accuracy99.2
16
Scene Text RecognitionStreet View Text 50-word lexicon (test)
Accuracy0.961
15
Scene Text RecognitionICDAR 2003 Full lexicon (test)
Accuracy98.6
11
Showing 10 of 13 rows

Other info

Code

Follow for update