Addressing Fisheries Policies in Pakistan by Using Deep Learning

Hinna Hafeez; Arqam Tariq; Zaib un Nisa; Farah Taj

doi:10.32350/jppp.31.01

Hinna Hafeez Superior University, Lahore, Pakistan
Arqam Tariq Government of the Punjab, Pakistan
Zaib un Nisa Superior University, Lahore, Pakistan
Farah Taj Superior University, Lahore, Pakistan

DOI: https://doi.org/10.32350/jppp.31.01

Keywords: convolutional Neural Network, Machine Learning Algorithm, features extraction, K- Nearest Neighbor, Machine learning algorithm, Support Vector Machine, Machine Learning Algorithm

Abstract

Abstract Views: 0

Pakistan has been facing an acute decline in the aquamarine population, especially fisheries, over the last few years. Therefore, the current study attempted to facilitate policymakers in identifying and addressing fish shortage using Artificial Intelligence (AI), particularly deep learning (DL). Policymakers can achieve this objective with the help of biologists and experts working in the field of AI. Presently, manual identification methods are still utilized in fish breeding, however, aquaculture is the only sector where the use of DL is increasing rapidly. For this, marine biologists and ichthyologists must have a precise taxonomy of fish species, if they intend to understand fish behavior deeply. Most current algorithms are designed to recognize fish in dry environments due to various challenges, such as background noise, picture distortion, the existence of other water bodies in images, low image quality, and occlusion. Due to the rapid growth of DL, the use of computer vision in agriculture and farming to generate agricultural intelligence has become a current hotspot in the field of research. Automatic classification of fish is severely limited by the inability to reliably distinguish between different fish species and taxonomic groups. Once the data is split into a "train" and "test" set, the former's features may be retrieved. Layer-specific feature extraction was performed for the current study. Subsequently, the model was trained using AlexNet and several machine-learning classification methods were compared to improve the classification accuracy. To demonstrate the use of DL in order to address the extinction of fisheries in Pakistan, a dataset of different types of fish was used, taken from Kaggle which measured 541MB. After choosing the dataset, AlexNet was used for the classification and to split the data into test (70%) and train data (30%). Afterwards, the features of train data were extracted on layers and AlexNet was used to train the model. Later, different machine learning classification algorithms were used to find the best classification accuracy, which may help to identify the fish breeds that are facing the threat of extinction. Moreover, policymakers may use the results to formulate policies in order to address the problem.

Downloads

Download data is not yet available.

References

Alaba, S. Y., Nabi, M. M., Shah, C., Prior, J., Campbell, M. D., Wallace, F., Ball, J. E., & Moorhead, R. (2022). Class-Aware fish species recognition using deep learning for an imbalanced dataset. Sensors, 22(21), Article e8268. https://doi.org/10.3390/s22218268

Carrington, A. M., Manuel, D. G., Fieguth, P. W., Ramsay, T., Osmani, V., Wernly, B., Bennett, C., Hawken, S., McInnes, M., Magwood, O., Sheikh, Y., & Holzinger, A. (2021). Deep ROC analysis and AUC as balanced average accuracy to improve model selection, understanding and interpretation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(1), 329–341. https://doi.org/10.1109/TPAMI.2022.3145392

Chen, L., Hua, H., Luo, X., Xu, G., & Yan, X. (2024). Prediction of the Economic Behavior of Fishery Biotechnology Companies Based on Machine Learning-Based Deep Metacellular Automata. arXiv preprint arXiv:2402.13509.

Hadlee, R. R., Gudadhe, A. A., Kumar, J. P., Puri, C. G., & Verma, P. (2023, November). An Analysis of Training Artificial Intelligence Techniques into Eco Sounder Machine to Identify Fish. In 2023 1st DMIHER International Conference on Artificial Intelligence in Education and Industry 4.0 (IDICAIEI) (Vol. 1, pp. 1-5). IEEE.

Hasegawa, T., & Nakano, D. (2024). Robust Fish Recognition Using Foundation Models toward Automatic Fish Resource Management. Journal of Marine Science and Engineering, 12(3), 488.

Iqbal, U., Li, D., & Akhter, M. (2022). Intelligent diagnosis of fish behavior using deep learning method. Fishes, 7(4), Article e201. https://doi.org/10.3390/fishes7040201

Ji, W., Peng, J., Xu, B., & Zhang, T. (2023). Real-time detection of underwater river crab based on multi-scale pyramid fusion image enhancement and MobileCenterNet model. Computers and Electronics in Agriculture, 204, Article e107522. https://doi.org/10.1016/j.compag.2022.107522

Jose, J. A., Kumar, C. S., & Sureshkumar, S. (2022, February). Automated region split algorithm with pre-trained CNN model for tuna classification. In 2022 First International Conference on Electrical, Electronics, Information and Communication Technologies (ICEEICT) (pp. 1-5). IEEE. https://doi.org/10.1109/ICEEICT53079.2022.9768589

Kandimalla, V., Richard, M., Smith, F., Quirion, J., Torgo, L., & Whidden, C. (2022). Automated detection, classification and counting of fish in fish passages with deep learning. Frontiers in Marine Science, 8, Article e823173. https://doi.org/10.3389/fmars.2021.823173

Khai, T. H., Abdullah, S. N. H. S., Hasan, M. K., & Tarmizi, A. (2022). Underwater fish detection and counting using mask regional convolutional neural network. Water, 14(2), Article e222. https://doi.org/10.3390/w14020222

Kumar, A. (2023, December 4). Different types of CNN architectures explained: Examples. Analytics Yogi. https://vitalflux.com/different-types-of-cnn-architectures-explained-examples/

Li, D., Su, H., Jiang, K., Liu, D., & Duan, X. (2022). Fish face identification based on rotated object detection: Dataset and exploration. Fishes, 7(5), Article e219. https://doi.org/10.3390/fishes7050219

MathWorks. (2024a). Deep Learning Toolbox: Design, train, analyze, and simulate deep learning networks. https://www.mathworks.com/help/deeplearning/ref/alexnet.html

MathWorks. (2024b). Resnet50: ResNet-50 convolutional neural network. https://www.mathworks.com/help/deeplearning/ref/resnet50.html

Mohsin, M., Yin, H., & Mehak, A. (2024). Sustainable solutions: Exploring risks and strategies in Pakistan’s seafood trade for marine conservation. Frontiers in Marine Science, 11, Article e1420755. https://doi.org/10.3389/fmars.2024.1420755

Nusrat, S. (2021). Fisheries: Potential of Pakistan. Trade Development Authority Pakistan. https://tdap.gov.pk/wp-content/uploads/2022/03/Fisheries-Potential-of-Pakistan-Salma-Nusrat.pdf

Pakistan’s seafood exports have massive growth potential (2023, August 31). The Nation. https://www.nation.com.pk/31-Aug-2023/pakistan-s-seafood-exports-have-massive-growth-potential

Rathi, D., Jain, S., & Indu, S. (2017, December 27–30). Underwater fish species classification using convolutional neural network and deep learning [Paper presentation]. In Proceedings of 2017 9th international conference on advances in pattern recognition. Bangalore, India.

Salman, A., Jalal, A., Shafait, F., Mian, A., Shortis, M., Seager, J., & Harvey, E. (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

Tamou, A. B., Benzinou, A., & Nasreddine, K. (2022a). Live fish species classification in underwater images by using convolutional neural networks based on incremental learning with knowledge distillation loss. Machine Learning and Knowledge Extraction, 4(3), 753–767. https://doi.org/10.3390/make4030036

Tamou, A. B., Benzinou, A., & Nasreddine, K. (2022b). Targeted data augmentation and hierarchical classification with deep learning for fish species identification in underwater images. Journal of Imaging, 8(8), Article e214. https://doi.org/10.3390/jimaging8080214