Abstract:
|
In recent years, text recognition has achieved remarkable success in recognizing scanned
document text. However, word recognition in natural images is still an open problem,
which generally requires time consuming post-processing steps. We present a novel architecture
for individual word detection in scene images based on semantic segmentation.
Our contributions are twofold: the concept of WordFence, which detects border areas
surrounding each individual word and a unique pixelwise weighted softmax loss function
which penalizes background and emphasizes small text regions. WordFence ensures that
each word is detected individually, and the new loss function provides a strong training
signal to both text and word border localization. The proposed technique avoids intensive
post-processing by combining semantic word segmentation with a voting scheme
for merging segmentations of multiple scales, producing an end-to-end word detection
system. We achieve superior localization recall on common benchmark datasets - 92%
recall on ICDAR11 and ICDAR13 and 63% recall on SVT. Furthermore, end-to-end
word recognition achieves state-of-the-art 86% F-Score on ICDAR13. |