Paper Title
Development of Web Application using Deep Learning for Classification of Fish Species

Abstract
Fish classification has emerged as a significant aspect of the development of agriculture, greatly influencing the aquaculture sector. Fish recognition is a labor-intensive, expensive process that necessitates a lot of manual labor, but artificial intelligence in the aquaculture field is making fish detection techniques simpler for researchers, marine biologists, etc. that can increase the production of aquaculture income for the economy. In order to avoid problems with fish detection, classification, and manual fish feeding, monitoring fish health, and disease detection, artificial intelligence is developing effective approaches for consistent monitoring. There are more than 22,000 species of fish known to exist, making it laborious and time-consuming to manually classify and categorize fishes. A contribution to contemporary methods of fish farming is the automatic detection of fishes, their count, image processing, and size with the use of various algorithms and coding applications of artificial intelligence models. The development of an automated classification system that can categorize fishes is the main topic of the present paper. By utilizing different transfer learning models of CNN, fish images give a testing accuracy of 99% on both the Large-Scale dataset and Fish-Pak Dataset. Keywords – Component; Freshwater Fish; Fish Species Classification; AlexNet; MobileNet; ResNet;