Movie Genre Classification through Movie posters using Deep Learning techniques
Neetu Singh, Rajput, Rahul, Muskan Jaggi, Bhatia, Rohan · 2023
Abstract: The classification of movie genres is a crucial task in the film industry for various purposes, such as recommendation systems, content analysis, and marketing strategies. Traditional methods primarily rely on textual features and metadata, which may not capture the visual essence of a movie accurately. In this research paper, we propose an IMDb genre classifier that utilizes movie posters as input through deep learning techniques. We demonstrate the effectiveness of our approach by training and evaluating a deep neural network on a large-scale dataset of movie posters and associated genres from IMDb. Our results show that leveraging deep learning techniques with movie posters significantly improves the genre classification performance, outperforming traditional text-based approaches. To implement the procedure with the maximum level of accuracy feasible, researchers in the domains of natural language processing (NLP) and machine learning (ML) have investigated several techniques. In this study, the feelings of the IMDb movie reviews are analysed. Effective preprocessing and partitioning of the data improved post-classification performance. The correctness of the classification performance is investigated. The best classification accuracy measured by the results is 89.9%. It demonstrates the viability of including the created solution in contemporary text-based sentiment analysers.