Sparse arrays of signatures for online character recognition

By Benjamin Graham. 2013

106 citations recorded by the source corpus.

In mathematics the signature of a path is a collection of iterated integrals, commonly used for solving dierential equations. We show that the path signature, used as a set of features for consumption by a convolutional neural network (CNN), improves the accuracy of online character recognition|that is the task of reading characters represented as a collection of paths. Using datasets of letters, numbers, Assamese and Chinese characters, we show that the rst, second, and even the third iterated integrals contain useful information for consumption by a CNN. On the CASIA-OLHWDB1.1 3755 Chinese character dataset, our approach gave a test error of 3.58%, compared with 5.61%[4] for a traditional CNN. A CNN trained on the CASIA-OLHWDB1.0-1.2 datasets won the ICDAR2013 Online Isolated Chinese Character recognition competition. Computationally, we have developed a sparse CNN implementation that make it practical to train CNNs with many layers of maxpooling. Extending the MNIST dataset by translations, our sparse CNN gets a test error of 0.31%.

Subfields: Computer Vision and Pattern Recognition, Statistical and Nonlinear Physics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

  1. Chenglin Liu
  2. Xu-Yao Zhang
  3. Fei Yin
  4. Yoshua Bengio
  5. Xiang Bai
  6. Luc Van Gool
  7. Dacheng Tao
  8. Cordelia Schmid

In Statistical and Nonlinear Physics

  1. Fei Yin
  2. Chenglin Liu
  3. Yoshua Bengio
  4. Yanming Zhang
  5. Xu-Yao Zhang
  6. Kyunghyun Cho
  7. Li Deng
  8. Hugo Larochelle

Citations in the corpus