Type-2 Fuzzy Deep Learning Framework for Robust Marine Species Identification Under Underwater Environmental Uncertainty
Keywords:
Deep learning, Karnik-Mendel type reduction, Marine species identification, Robust classification, Type-2 fuzzy logic, Underwater image uncertaintyAbstract
Underwater imagery is characterized by non-stationary optical noise, unpredictable turbidity, and inconsistent illumination that severely undermine the reliability of conventional deep learning classifiers used for marine species identification. This study proposes IT2F-AquaNet, an interval type-2 fuzzy deep learning framework that embeds a second layer of uncertainty modeling, termed the footprint of uncertainty, directly within a convolutional-transformer hybrid backbone to explicitly represent both aleatoric image noise and epistemic labeling ambiguity. The framework couples an adaptive preprocessing stage, an Interval Type-2 Fuzzy Logic System (IT2FLS) uncertainty module, a fuzzy feature-fusion layer employing Karnik-Mendel type reduction, and an uncertainty-aware attention classifier that outputs both a species label and a calibrated confidence band. The framework was evaluated on a composite benchmark built from multiple public underwater datasets augmented with simulated turbidity, illumination drift, and motion blur. Experimental results demonstrate that IT2F-AquaNet attains 93.8% classification accuracy and a macro F1-score of 92.9%, outperforming six state-of-the-art baselines by margins of 5.4 to 9.6 percentage points under degraded visual conditions. Ablation analysis confirms that the fuzzy uncertainty module contributes the largest performance gain among all architectural components. These findings position type-2 fuzzy-augmented deep architectures as a viable direction for dependable marine biodiversity monitoring systems.
References
J. S. Jaffe, K. D. Moore, J. McLean, and M. P. Strand, “Underwater Optical Imaging: Status and Prospects,” Oceanography, vol. 14, no. 3, pp. 64–75, 2001.
J. Zhou, T. Yang, W. Ren, D. Zhang, and W. Zhang, "Underwater image restoration via depth map and illumination estimation based on a single image," Optics Express, vol. 29, no. 19, pp. 29864-29886, 2021.
A. Saleh, I. H. Laradji, D. A. Konovalov, M. D. Bradley, D. Vazquez, and M. Sheaves, “A realistic fish-habitat dataset to evaluate algorithms for underwater visual analysis,” Scientific Reports, vol. 10, no. 1, p. 14671, Sept. 2020.
S. Villon et al., “A Deep learning method for accurate and fast identification of coral reef fishes in underwater images,” Ecological Informatics, vol. 48, pp. 238–244, Nov. 2018.
W. W. Tan and T. W. Chua, "Uncertain Rule-Based Fuzzy Logic Systems: Introduction and New Directions (Mendel, J.M.; 2001) [book review]," in IEEE Computational Intelligence Magazine, vol. 2, no. 1, pp. 72-73, Feb. 2007.
J. M. Mendel and R. I. B. John, "Type-2 fuzzy sets made simple," in IEEE Transactions on Fuzzy Systems, vol. 10, no. 2, pp. 117-127, April 2002.
O. Castillo and P. Melin, “A review on the design and optimization of interval type-2 fuzzy controllers,” Applied Soft Computing, vol. 12, no. 4, pp. 1267–1278, Apr. 2012.
M. Abdar et al., “A review of uncertainty quantification in deep learning: Techniques, applications and challenges,” Information Fusion, vol. 76, no. 76, pp. 243–297, Dec. 2021.
C. O. Ancuti, C. Ancuti, C. De Vleeschouwer, and P. Bekaert, “Color Balance and Fusion for Underwater Image Enhancement,” IEEE Transactions on Image Processing, vol. 27, no. 1, pp. 379–393, Jan. 2018.
C. Li, J. Guo and C. Guo, "Emerging From Water: Underwater Image Color Correction Based on Weakly Supervised Color Transfer," in IEEE Signal Processing Letters, vol. 25, no. 3, pp. 323-327, March 2018.
K. He, X. Zhang, S. Ren and J. Sun, "Deep Residual Learning for Image Recognition," 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 2016, pp. 770-778.
M. Tan and Q. Le, “EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks,” PMLR, pp. 6105–6114, May 2019.
A. Dosovitskiy et al., “An image is worth 16x16 words: Transformers for image recognition at scale,” arXiv preprint arXiv:2010.11929, 2020 Oct 22.
L. A. Zadeh, “Fuzzy sets,” Information and Control, vol. 8, no. 3, pp. 338–353, June 1965.
D. Hidalgo, P. Melin, and O. Castillo, “An optimization method for designing type-2 fuzzy inference systems based on the footprint of uncertainty using genetic algorithms,” Expert Systems with Applications, vol. 39, no. 4, pp. 4590–4598, Mar. 2012.
L. A. Zadeh, “The concept of a linguistic variable and its application to approximate reasoning—I,” Information Sciences, vol. 8, no. 3, pp. 199–249, 1975.
M. Yaqub et al., “State-of-the-Art CNN Optimizer for Brain Tumor Segmentation in Magnetic Resonance Images,” Brain Sciences, vol. 10, no. 7, p. 427, July 2020.
P. Ghamisi et al., "Multisource and Multitemporal Data Fusion in Remote Sensing: A Comprehensive Review of the State of the Art," in IEEE Geoscience and Remote Sensing Magazine, vol. 7, no. 1, pp. 6-39, March 2019.
C. Blundell, Julien Cornebise, Koray Kavukcuoglu, and Daan Wierstra, “Weight Uncertainty in Neural Network,” PMLR, pp. 1613–1622, June 2015.
S. Murat, L. Kaplan, and M. Kandemir, “Evidential Deep Learning to Quantify Classification Uncertainty,” Advances in Neural Information Processing Systems, vol. 31, 2018.
B. J. Boom, P. X. Huang, J. He and R. B. Fisher, "Supporting ground-truth annotation of image datasets using clustering," Proceedings of the 21st International Conference on Pattern Recognition (ICPR2012), Tsukuba, Japan, 2012, pp. 1542-1545.
M. J. Islam et al., "Semantic Segmentation of Underwater Imagery: Dataset and Benchmark," 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Las Vegas, NV, USA, 2020, pp. 1769-1776.
S. Woo, J. Park, J.-Y. Lee, and I. S. Kweon, “CBAM: Convolutional Block Attention Module,” Lecture Notes in Computer Science, 2018, pp. 3–19.
