Film Genre Classification Based on Poster Using Convolutional Neural Network

Abstract

This study proposes a method for classifying film genres based on their posters using Convolutional Neural Networks (CNNs). Film posters are valuable visual representations that encapsulate key elements of a movie's genre. The CNN model is trained on a dataset comprising diverse film posters and their corresponding genres. The model learns to extract relevant features from the posters and classify them into predefined genres. Experimental results demonstrate the effectiveness of the proposed approach in accurately classifying film genres solely based on their posters, showcasing the potential for automated genre classification systems in the film industry.