ShrimpXNet: AI-Powered Disease Detection for Sustainable Aquaculture

This research presents a practical application of transfer learning and adversarial training for a critical problem in aquaculture. While the results are promising, the relatively small dataset size (1,149 images) raises concerns about the generalizability of the model to diverse real-world conditions and unseen disease variations. Further validation with larger, more diverse datasets is crucial.

• •ShrimpXNet utilizes transfer learning for shrimp disease classification.
• •Adversarial training and data augmentation enhance model robustness.
• •ConvNeXt-Tiny achieved 96.88% accuracy in disease classification.