Exploiting Fisher kernels in decoding severely noisy document images
Decoding noisy document images is commonly needed in applications such as enterprise content management. Available OCR solutions are still not satisfactory especially on noisy images, and re-trainable systems require difficult and tedious training example preparation. Motivated by this challenging real application, we propose a novel solution that organically combines generative OCR models with discriminative classification via a RBF Fisher kernel derived from an independent bit-flip template model. We show that the new approach is highly accurate in decoding noisy document images, making the system more generalizable to variations in font and degradation, and hence significantly reduces the burden in training example preparation. We also show that Fisher kernel can be used to reduce feature dimension and to build simpler and more robust models.
- download PDF (141K)
Chen, J. D. ; Wang, Y. Exploiting Fisher kernels in decoding severely noisy document images. Ninth International Conference on Document Analysis and Recognition (ICDAR 2007); 2007 September 23-26; Curitiba; Brazil.
Copyright © IEEE, 2007. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.