Research on Image Recognition of Marine Organisms Based on MIRNet and ViT Models

Xi Chen; Yuxi Luo; Wenjing Xu; Jiayi Zhao

doi:10.2991/978-94-6463-512-6_62

<Previous Article In Volume

Next Article In Volume>

Research on Image Recognition of Marine Organisms Based on MIRNet and ViT Models

Authors

Xi Chen¹, Yuxi Luo², Wenjing Xu³, Jiayi Zhao⁴^{, *}

¹FEIT, University of Melbourne, Melbourne, VIC, 3052, Australia

²School of Computing and Data Science, Xiamen University Malaysia, Sepang, Selangor, 43900, Malaysia

³Software College, Jiangxi Agricultural University, Nanchang, Jiangxi Province, 330000, China

⁴Software Engineer, Dalian University of Technology, Dalian, Liaoning Province, 116000, China

^*Corresponding author. Email: 20212251035@mail.dlut.edu.cn

Corresponding Author

Jiayi Zhao

Available Online 23 September 2024.

DOI: 10.2991/978-94-6463-512-6_62 How to use a DOI?
Keywords: Image Enhancement; Transfer Learning; ViT
Abstract: With the deterioration of the marine environment, it is crucial to protect marine biodiversity. This paper implements and compares the performance of various models for their accuracy in species identification. ResNet50V2 obtained an accuracy of 79.62% on its validation set, according to the study's findings. 78.79% accuracy was attained by MobileNetV2 on its validation set. On the verification set, EfficientNetB7 achieved an accuracy of 73.57%. These models’ performance was not as good as ViT's, which beat the competition with reduced loss rates and an accuracy of 91.51% on the training set and 90.23% on the validation set. Ultimately, the study sought to improve overall identification accuracy by improving low-quality photos. Subsequent studies using the MIRNetV2 model yielded greater results than the MIRNet model; it demonstrated strong image improvement capabilities, achieving an accuracy of 90.34% on the training set and 89.84% on the validation set. The findings suggest that improving image quality significantly enhances species identification accuracy, contributing to preserving marine biodiversity.
Open Access: Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

<Previous Article In Volume

Next Article In Volume>

Volume Title: Proceedings of the 2024 International Conference on Artificial Intelligence and Communication (ICAIC 2024)
Series: Advances in Intelligent Systems Research
Publication Date: 23 September 2024
ISBN: 978-94-6463-512-6
ISSN: 1951-6851
DOI: 10.2991/978-94-6463-512-6_62 How to use a DOI?
Open Access: Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Cite this article

ris enw bib

TY  - CONF
AU  - Xi Chen
AU  - Yuxi Luo
AU  - Wenjing Xu
AU  - Jiayi Zhao
PY  - 2024
DA  - 2024/09/23
TI  - Research on Image Recognition of Marine Organisms Based on MIRNet and ViT Models
BT  - Proceedings of the 2024 International Conference on Artificial Intelligence and Communication (ICAIC 2024)
PB  - Atlantis Press
SP  - 586
EP  - 599
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-512-6_62
DO  - 10.2991/978-94-6463-512-6_62
ID  - Chen2024
ER  -

download .riscopy to clipboard