Improved deep learning image classification algorithm based on Swin Transformer V2
PeerJ Computer Science · 30.10.2023 · 9 · e1665
DOI: 10.7717/peerj-cs.1665 ↗Kurz gefasst
Fachartikel in der Zeitschrift PeerJ Computer Science, online erschienen am 30.10.2023. Verfasst von Jiangshu Wei und 4 weiteren Autor:innen. Beteiligte Einrichtung: College of Information Engineering, Sichuan Agricultural University, Ya'an, Sichuan, China. Die Arbeit wurde bisher 13-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: University-Industry Collaborative Education Program.
Abstract (Originalsprache)
While convolutional operation effectively extracts local features, their limited receptive fields make it challenging to capture global dependencies. Transformer, on the other hand, excels at global modeling and effectively captures global dependencies. However, the self-attention mechanism used in Transformers lacks a local mechanism for information exchange within specific regions. This article attempts to leverage the strengths of both Transformers and convolutional neural networks (CNNs) to enhance the Swin Transformer V2 model. By incorporating both convolutional operation and self-attention mechanism, the enhanced model combines the local information-capturing capability of CNNs and the long-range dependency-capturing ability of Transformers. The improved model enhances the extraction of local information through the introduction of the Swin Transformer Stem, inverted residual feed-forward network, and Dual-Branch Downsampling structure. Subsequently, it models global dependencies using the improved self-attention mechanism. Additionally, downsampling is applied to the attention mechanism’s Q and K to reduce computational and memory overhead. Under identical training conditions, the proposed method significantly improves classification accuracy on multiple image classification datasets, showcasing more robust generalization capabilities.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- College of Information Engineering, Sichuan Agricultural University, Ya'an, Sichuan, China.
- College of Mechanical and Electrical Engineering, Sichuan Agricultural University, Ya'an, Sichuan, China.
Förderung
- University-Industry Collaborative Education Program
Zitieren
APA
Wei, J., Chen, J., Wang, Y., Luo, H., & Li, W. (2023). Improved deep learning image classification algorithm based on Swin Transformer V2. PeerJ Computer Science, 9, e1665. https://doi.org/10.7717/peerj-cs.1665
Vancouver
Wei J, Chen J, Wang Y, Luo H, Li W. Improved deep learning image classification algorithm based on Swin Transformer V2. PeerJ Computer Science. 2023;9:e1665. doi:10.7717/peerj-cs.1665
BibTeX
@article{Wei2023,
author = {Wei, Jiangshu and Chen, Jinrong and Wang, Yuchao and Luo, Hao and Li, Wujie},
title = {Improved deep learning image classification algorithm based on Swin Transformer V2},
journal = {PeerJ Computer Science},
year = {2023},
volume = {9},
pages = {e1665},
doi = {10.7717/peerj-cs.1665}
}
EUVIMED