Beyond human recognition: A CNN-based framework for handwritten character recognition
Authors:
- 1 November 2015
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 695-699
Abstract
Because of the various appearance (different writers, writing styles, noise, etc.), the handwritten character recognition is one of the most challenging task in pattern recognition. Through decades of research, the traditional method has reached its limit while the emergence of deep learning provides a new way to break this limit. In this paper, a CNN-based handwritten character recognition framework is proposed. In this framework, proper sample generation, training scheme and CNN network structure are employed according to the properties of handwritten characters. In the experiments, the proposed framework performed even better than human on handwritten digit (MNIST) and Chinese character (CASIA) recognition. The advantage of this framework is proved by these experimental results.Keywords
This publication has 10 references indexed in Scilit:
- DeepFace: Closing the Gap to Human-Level Performance in Face Verification2014
- ICDAR 2013 Chinese Handwriting Recognition Competition2013
- Multi-column deep neural network for traffic sign classificationNeural Networks, 2012
- Multi-column deep neural networks for image classification2012
- Convolutional Neural Network Committees for Handwritten Character Classification2011
- Normalization-Cooperated Gradient Feature Extraction for Handwritten Character RecognitionIEEE Transactions on Pattern Analysis and Machine Intelligence, 2007
- Best practices for convolutional neural networks applied to visual document analysis2005
- Handwritten digit recognition: investigation of normalization and feature extraction techniquesPattern Recognition, 2004
- Affine-invariant recognition of gray-scale characters using global affine transformation correlationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2001
- Aspect Ratio Adaptive Normalization for Handwritten Character Recognition2000