WAVELET-BASED EFFECTIVE IMAGES DOWNSCALING IN NEURAL NETWORKS

Article Sidebar

Abstract views: 30
PDF Downloads: 6
pdf (rus) (Русский)
Published: Nov 18, 2024
DOI: https://doi.org/10.52928/2070-1624-2024-43-2-10-15
Keywords:
нейронные сети, глубокое обучение, вейвлеты, дискретное вейвлет-преобразование, классификация изображений neural networks, deep learning, wavelets, discrete wavelet transform, image classification
Issue
No. 2 (2024)
Section
Информатика, вычислительная техника и управление

Main Article Content

U. VARABEI
Belarusian State University, Minsk
https://orcid.org/0009-0006-9604-8894
A. MALEVICH
Belarusian State University, Minsk
https://orcid.org/0000-0001-8716-8655

Abstract

Based on discrete wavelet transform, several blocks for images downscaling in computer vision models were implemented. The blocks were tested with ResNetV2-50 and MobileNetV2 models on Flowers dataset. With small increase in number of models’ parameters and close results in terms of metric the changes allowed to reduce number of training epochs by 34 % and VRAM requirements by 18 %. Due to the implementation details the blocks suggested can be used as a replacement of layers responsible for images downscaling in models for other tasks to save computation resources and speed up training process. In the blocks developed standard operations of addition and multiplication are used for evaluation of wavelet transform, which allows a simple export of trained models into other formats.

Article Details

How to Cite
VARABEI, U., & MALEVICH, A. (2024). WAVELET-BASED EFFECTIVE IMAGES DOWNSCALING IN NEURAL NETWORKS. Vestnik of Polotsk State University. Part C. Fundamental Sciences, (2), 10-15. https://doi.org/10.52928/2070-1624-2024-43-2-10-15
Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Author Biography

A. MALEVICH, Belarusian State University, Minsk

канд. физ.-мат. наук, доц.

Crossref
Scopus
Google Scholar
Europe PMC

References

Huang, G., Liu, Z., van der Maaten, L., & Weinberger, K. Q. (2016). Densely Connected Convolutional Networks. arXiv. DOI: 10.48550/arXiv.1608.06993.

Hu, J., Shen, L., Albanie, S., Sun, G., & Wu, E. (2017). Squeeze-and-Excitation Networks. arXiv. DOI: 10.48550/arXiv.1709.01507.

Wang, C.-Y., Liao, H.-Y. M., Yeh, J-H., Wu, Y.-H., Chen, P.-Y., Hsieh, J.-W. (2019). CSPNet: A New Backbone than can Enhance Learning Capability of CNN. arXiv. DOI: 10.48550/arXiv.1911.11929.

Li, Q., Shen, L., Guo, S., Lai, Z. (2020). Wavelet Integrated CNNs for Noise-Robust Image Classification. arXiv. DOI: 10.48550/arXiv.2005.03337.

Liu, P., Zhang, H., Lian, W., & Zuo, W. (2019). Multi-level Wavelet Convolutional Neural Networks. arXiv. DOI: 10.48550/arXiv.1907.03128.

Xiangyu, Z. (2022). Wavelet-Attention CNN for Image Classification. arXiv. DOI: 10.48550/arXiv.2201.09271.

Woo, S., Park, J., Lee, J.-Y., & Kweon, I. S. (2018). CBAM: Convolutional Block Attention Module. arXiv. DOI: 10.48550/arXiv.1807.06521.

He, K., Zhang, X., Ren, S., Sun, J. (2016). Identity Mappings in Deep Residual Networks. arXiv. DOI: 10.48550/arXiv.1603.05027.

Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., Chen, L.-C. (2018). MobileNetV2: Interested Residuals and Linear Bottlenecks. arXiv. DOI: 10.48550/arXiv.1801.04381.

Getreuer, P. (1997). Wavelet CDF 9/7 Implementation. URL: https://getreuer.info/posts/waveletcdf97/index.html.