CNN training with graph-based sample preselection: application to handwritten character recognition

2017/12/06 by Frédéric Rayar, Rayar, Frédéric, Masanori Goto +3
Computer Science · #Advanced Neural Network Applications #FOS: Computer and information sciences #Face and Expression Recognition #Handwritten Text Recognition Techniques #Machine Learning (cs.LG)

paper · pdf · doi:10.48550/arxiv.1712.02122

openalex publication_date 2017/12/06 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In this paper, we present a study on sample preselection in large training data set for CNN-based classification. To do so, we structure the input data set in a network representation, namely the Relative Neighbourhood Graph, and then extract some vectors of interest. The proposed preselection method is evaluated in the context of handwritten character recognition, by using two data sets, up to several hundred thousands of images. It is shown that the graph-based preselection can reduce the training data set without degrading the recognition accuracy of a non pretrained CNN shallow model.

Citations

• Asymptotic Properties of Nearest Neighbor Rules Using Edited Data
• Gradient-based learning applied to document recognition

Related

• Text recognition on images using pre-trained CNN
• Assessing four Neural Networks on Handwritten Digit Recognition Dataset (MNIST)
• Spatially-sparse convolutional neural networks
• Gradient-based learning applied to document recognition
• Convolutional Neural Network with Pruning Method for Handwritten Digit Recognition