Classification of Seashells Using Deep Learning Models

Seashell classification is essential for industries, including food, jewelry, decor, pharmaceuticals, and tourism. Conventional classification techniques mainly rely on humans, which is error-prone and time-consuming, requiring manual expertise. This paper uses a deep learning-based approach for classifying seashells into four categories. Our method employs deep learning models, including SENet and InceptionNet-V3. Our study reveals that SENet achieved a good accuracy of 99.46% at a train-test ratio of 80:20 with a batch size of 8 and 60 epochs.