Noise-Based and Class-Based Curriculum Learning for Image Classifiers

2023 15th International Conference on Computer Research and Development (ICCRD 2023): Proceedings 2023
Yue Li, Ēvalds Urtāns

Datasets often contain different difficulty samples and even noisy samples. This paper introduces two naive curriculum learning methods, one using an image dataset with noise and another one using an image dataset that contains samples from other datasets with presumed higher difficulty. The final goal is to improve the performance of the model by gradually introducing more difficult samples during the training process rather than using them from the very beginning. Experiments demonstrated that using the proposed curriculum learning methods, a classifier can achieve higher accuracy in less training epochs.

Atslēgas vārdi
Deep Learning, Curriculum Learning, Image Classification
DOI
10.1109/ICCRD56364.2023.10080241
Hipersaite
https://ieeexplore.ieee.org/document/10080241

Li, Y., Urtāns, Ē. Noise-Based and Class-Based Curriculum Learning for Image Classifiers. No: 2023 15th International Conference on Computer Research and Development (ICCRD 2023): Proceedings, Ķīna, Hangzhou, 10.-12. janvāris, 2023. Piscataway: IEEE, 2023, 161.-166.lpp. ISBN 978-1-6654-8751-1. e-ISBN 978-1-6654-8750-4. ISSN 2161-0886. e-ISSN 2161-0894. Pieejams: doi:10.1109/ICCRD56364.2023.10080241

Publikācijas valoda
English (en)