MNIST handwritten digit recognition with different CNN architectures
Journal of Applied Technology and Innovation · 01.03.2021 · 5 (1)
DOI: 10.65136/jati.v5i1.195 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Applied Technology and Innovation, online erschienen am 01.03.2021. Verfasst von Lead Ming Seng und 4 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
Handwritten digit recognition has long been a popular research topic in computer vision and pattern recognition. Recognizing handwritten digits used to be challenging but thanks to many machine learning techniques nowadays, the problem is no longer. In this research, we looked into the MNIST database using fast.ai and trained the CNN ResNet-18 model to recognize handwritten digits. We then modified the architecture with different pre-trained models. For this work, we implemented five PyTorch’s pre-trained models, which are GoogLeNet, MobileNet v2, ResNet-50, ResNeXt-50, Wide ResNet-50. The purpose of this paper is to reveal the most accurate architecture for handwritten digits recognition. Also, we provide comparisons of training time, top-1 error, top-5 error and model size on all five models.
Zitieren
APA
Seng, L. M., Chiang, B. B. C., Salam, Z. A. A., Tan, G. Y., & Chai, H. T. (2021). MNIST handwritten digit recognition with different CNN architectures. Journal of Applied Technology and Innovation, 5(1). https://doi.org/10.65136/jati.v5i1.195
Vancouver
Seng LM, Chiang BBC, Salam ZAA, Tan GY, Chai HT. MNIST handwritten digit recognition with different CNN architectures. Journal of Applied Technology and Innovation. 2021;5(1). doi:10.65136/jati.v5i1.195
BibTeX
@article{Seng2021,
author = {Seng, Lead Ming and Chiang, Brennan Bang Chen and Salam, Zailan Arabee Abdul and Tan, Gwo Yih and Chai, Hui Tong},
title = {MNIST handwritten digit recognition with different CNN architectures},
journal = {Journal of Applied Technology and Innovation},
year = {2021},
volume = {5},
number = {1},
doi = {10.65136/jati.v5i1.195}
}
EUVIMED