Milkfish, a dietary staple for many, is highly ResNet50 architecture to classify milkfish into fresh and susceptible to spoilage. Maintaining its freshness is not - fresh categories, (Mahendran, R., & Seneviratne, G. crucial to preserve both its quality and nutritional value. P. (2022). Their model achieved promising results, While traditional methods like icing have been widely demonstrating the potential of deep learning for accurate used, they often rely on subjective asses sments. and efficient freshness assessmen t. Another study by In recent years, image processing techniques Abu Rayan et al. (2021) utilized a combined deep have emerged as a promising solution to objectively learning model, incorporating VGG - 16 and Bi - directional Long Short - Term Memory (BiLSTM) assess fish freshness. By analyzing specific features like eye clarity, gill color, and tissue integrity, artificial networks, to classify fish freshness with high accuracy. intelligence models, such as Convolutional Neural Networks (CNNs), can accurately determine whether a By developing and deploying deep learning milkfish is fresh or not. models, it is possible to automate the process of fish Our research focused on developing a CNN freshness classification, Aguil, A. J. C., et.al. (2018). model, to analyze images of milkfish. By segmenting reducing the reliance on human judgment and key areas like the eyes and body tissues, the model can improving the overall quality control in the seafood extract relevant features and make precise predictions. indu stry. This technology can help ensure that