A Transfer Learning-Based Approach for Classification of Fish Species Using Machine and Deep Learning Techniques.
Keywords:
Marine fish classification; Deep learning; Transfer learning; ResNet50; VGG16; InceptionV3; MobileNetV2; Image augmentation; Grad-CAM; Biodiversity monitoring; Sustainable fishery managementAbstract
Precise identification of marine fish species is essential to promote sustainable fishery management and preserve marine biodiversity. This research explores the performance of four pre-trained deep learning architectures VGG16, ResNet50, InceptionV3, and MobileNetV2 in a transfer learning paradigm to identify nine marine fish species. A balanced and augmented dataset of 9,000 images (1,000 per species) was employed to promote model generalization and robustness. All the models performed excellently, with ResNet50 recording the highest classification accuracy of 99.72%, followed closely by VGG16 (99.70%), InceptionV3 (99.60%), and MobileNetV2 (99.44%). While overall performance was great, there were minor misclassifications between visually similar species. Model interpretability was enhanced with the Grad-CAM technique, demonstrating that the models effectively attended to relevant morphological features during prediction. Comparing previous work shows that the suggested approach performs better than state-of-the-art models at accuracy even when it is identifying more species. The findings demonstrate the promise of state-of-the-art deep learning for supporting automated, scalable, and accurate marine species classification for environmental and ecological applications.
Downloads
References
[1]. G. Kaur, N. Adhikari, S. Krishnapriya, S. G. Wawale, R. Q. Malik, A. S. Zamani, J. Perez-Falcon, and J. Osei-Owusu, "Recent Advancements in Deep Learning Frameworks for Precision Fish Farming: Opportunities, Challenges, and Applications," Journal of Food Quality, vol. 2023, Article ID 4399512, 2023. Available: https://doi.org/10.1155/2023/4399512
[2] Zhou, C., Li, H., Xie, Y., & Xu, M. (2022). Integrating taxonomic resolution into fisheries assessment: A framework for sustainable marine biodiversity management. Marine Policy, 140, 105051. https://doi.org/10.1016/j.marpol.2022.105051
[3] Qin, Y., Wang, Y., & Zhang, Q. (2023). Advancing conservation efforts through high-resolution species classification in aquatic ecosystems. Aquatic Conservation: Marine and Freshwater Ecosystems, 33(1), 25–38. https://doi.org/10.1002/aqc.3920
[4] Pitcher, T. J. (2008). The Sea Ahead: Challenges to Marine Biology in the 21st Century. Hydrobiologia, 606(1), 1–11.
[5] Ali-Gombe, A., Wang, Y., & Wang, Y. (2017). Fish Species Classification Using Convolutional Neural Network. In 2017 IEEE International Conference on Imaging Systems and Techniques (IST) (pp. 1–6). https://doi.org/10.1109/IST.2017.8261466
[6] Alsmadi, M. K., & Almarashdeh, I. A. (2022). A Review on Fish Classification Techniques in Underwater Environments. Journal of King Saud University - Computer and Information Sciences, 34(8), 5123–5134. https://doi.org/10.1016/j.jksuci.2020.10.004
[7] Chen, Z., Wang, Y., Lin, T., & Xu, H. (2023). Deep learning-driven segmentation for aquatic species recognition: A scalable alternative to manual taxonomic analysis. Ecological Informatics, 75, 102209. https://doi.org/10.1016/j.ecoinf.2023.102209
[8] Batista, G. E. A. P. A., Prati, R. C., & Monard, M. C. (2004). A study of the behavior of several methods for balancing machine learning training data. ACM SIGKDD Explorations Newsletter, 6(1), 20–29. https://doi.org/10.1145/1007730.1007735
[9] Alaba, F. A., Othman, M., Hashem, I. A. T., & Yau, K. L. A. (2022). Machine learning applications in aquatic ecosystem monitoring: A review. Environmental Monitoring and Assessment, 194(3), 1–25. https://doi.org/10.1007/s10661-021-09691-9
[10] Cabreira, Á. G., Andrade, A. F. A., & Marques, M. C. M. (2009). Using machine learning to predict fish species richness in streams. Ecological Informatics, 4(2), 117–124. https://doi.org/10.1016/j.ecoinf.2009.01.003
[11] Aguado, E., García, M., & Zúñiga, J. (2016). Fish species classification using machine learning techniques.
Ecological Informatics, 33, 1–7. https://doi.org/10.1016/j.ecoinf.2016.03.003
[12] Salman, A., Jalal, A., Shamsher, M., & Mian, A. (2016). Fish species classification in unconstrained underwater environments based on deep learning. Limnology and Oceanography: Methods, 14(9), 570–585. https://doi.org/10.1002/lom3.10113
[13] Feldmann, A., Hasan, M. K., & Peters, R. (2021). Deep learning for underwater fish species classification using convolutional neural networks. Sensors, 21(9), 3020. https://doi.org/10.3390/s21093020
[14] Sun, X., Yu, Y., Song, W., Zhang, Y., & Wang, J. (2017). Fish species classification by using convolutional neural network and generative adversarial networks. Proceedings of the 2017 International Conference on Image, Vision and Computing (ICIVC), 783–787. https://doi.org/10.1109/ICIVC.2017.7984637
[15] Zafar, U., Khan, A., Rho, S., Ahmed, E., & Ko, M. (2022). Deep learning-based fish classification model using convolutional neural networks. Computers and Electronics in Agriculture, 192, 106610. https://doi.org/10.1016/j.compag.2021.106610
[16] Cristianini, N., & Shawe-Taylor, J. (2000). An Introduction to Support Vector Machines and Other Kernel-based Learning Methods. Cambridge University Press. https://doi.org/10.1017/CBO9780511801389
[17] Zhang, J., Chen, Y., Li, D., & Huang, Y. (2017). Fish species classification based on K-nearest neighbor algorithm. Journal of Intelligent & Fuzzy Systems, 32(5), 3453–3460. https://doi.org/10.3233/JIFS-169651
[18] A. Singh, D. S. Chauhan, and A. Choubey, "Pattern Recognition Using Gaussian Mixture Model," International Journal of Computer Applications, vol. 3, no. 5, pp. 39–43, June 2010, doi: 10.5120/709-998.https://doi.org/10.5120/709-998
[19] C. Tan, F. Sun, T. Kong, W. Zhang, C. Yang, and C. Liu, "A Survey on Deep Transfer Learning," in Proceedings of the 27th International Conference on Artificial Neural Networks (ICANN), Rhodes, Greece, 2018, pp. 270–279. [Online]. Available: https://doi.org/10.1007/978-3-030-01424-7_27
[20] M. E. Taylor and P. Stone, "Transfer Learning for Reinforcement Learning Domains: A Survey," Journal of Machine Learning Research, vol. 10, pp. 1633–1685, 2009. [Online]. Available: http://www.jmlr.org/papers/volume10/taylor09a/taylor09a.pdf
[21] Thenmozhi, K., & Reddy, U. S. (2019). Crop pest classification based on deep convolutional neural network and transfer learning. Computers and Electronics in Agriculture, 164, 104906. https://doi.org/10.1007/s11042-023-15221-3
