• Home
  • Search
  • CNN Training with Graph-Based Sample Preselection: Application to Handwritten Character Recognition
  • Open Access IconOpen Access
  • Cite Icon9
  • https://doi.org/10.1109/das.2018.10Copy DOI Icon

CNN Training with Graph-Based Sample Preselection: Application to Handwritten Character Recognition

  • Apr 1, 2018
  • Frederic Rayar +2 more
Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
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.

Similar Papers
  • Research Article
  • Citations1013

The relative neighbourhood graph of a finite planar set

  • Jan 01, 1980
  • Pattern Recognition
  • Godfried T Toussaint
  • PDF
  • Research Article
  • Citations56

Pushing the limits of solubility prediction via quality-oriented data selection.

  • Dec 17, 2020
  • iScience
  • Murat Cihan Sorkun +2
  • Conference Article

Analysis of the energy consumption for the clustered topology constructed by combining RNG and GCA

  • Jun 01, 2012
  • Chih-Cheng Tseng +5
  • Research Article
  • Citations2

(Digital Presentation) Thermal Behaviour Prediction of Commercial Lithium-Ion Cells Under Different C-Rate and Ambient Conditions Using Surrogate Modelling

  • Jul 07, 2022
  • Electrochemical Society Meeting Abstracts
  • Raghvendra Gupta +3
  • Research Article

Sharp thresholds for relative neighborhood graphs in wireless ad hoc networks

  • Feb 01, 2010
  • IEEE Transactions on Wireless Communications
  • Yichih-Wei +3
  • Front Matter
  • Citations45

Signal quality in cardiorespiratory monitoring

  • Aug 17, 2012
  • Physiological Measurement
  • Gari D Clifford +1
  • Conference Article
  • Citations5

Combination of Relative Neighborhood Graph and Forbidden Set in the Design of Distributed Broadcast Algorithms for Wireless Ad Hoc Networks

  • Oct 01, 2010
  • Hwang-Cheng Wang +3
  • Research Article
  • Citations9

Comparative analysis of deep convolution neural network models on small scale datasets

  • Nov 13, 2022
  • Optik
  • C.R Edwin Selva Rex +2
  • Conference Article
  • Citations13

Classification Of a bank data set on various data mining platforms

  • Apr 01, 2018
  • Muhammet Sinan Basarslan +1
  • Research Article
  • Citations52

The effects of data preprocessing in the determination of coagulant dosing rate

  • Jul 27, 2000
  • Water Research
  • Dae-Sung Joo +2
  • Research Article
  • Citations12

DNAinsight: An Image Processing System for 2-D Gel Electrophoreisis of Genomic DNA.

  • Jan 01, 1997
  • Genome Informatics
  • Watanabe +2
  • Research Article
  • Citations8

IFN-signaling gene expression as a diagnostic biomarker for monogenic interferonopathies

  • Jun 17, 2024
  • JCI Insight
  • Laura A Adang +10
  • Abstract
  • Citations11

Three Biomarker Panel at Day 7 and 14 Can Predict Development of Grade II-IV Acute Graft-Versus-Host Disease

  • Nov 19, 2010
  • Blood
  • Sophie Paczesny +19
  • Research Article
  • Citations3

Revisiting relative neighborhood graph-based broadcasting algorithms for multimedia ad hoc wireless networks

  • Aug 02, 2011
  • The Journal of Supercomputing
  • Hwang-Cheng Wang +4
  • Conference Article
  • Citations5

DGG, dRNG, DSC: New Degree-based Shape-based Faithfulness Metrics for Large and Complex Graph Visualization

  • Apr 01, 2022
  • Seok-Hee Hong +5
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.