Movie Genre Classification Based on Poster and Subtitles Using Hybrid Combination of Convolutional Neural Networks

Automatic genre detection in movies is an important and catchy topic that can be used in many applications and contexts by different industries, such as personal development systems, database management, content analysis systems, and marketing and advertising systems. There are still a few gaps that have yet to be covered by research in the arena of movie genre detection. To address such challenges, we have proposed the hybrid methodology of applying image processing of posters along with subtitle text in order to identify movie genres automatically. It employs Convolutional Neural Networks (CNN) with preprocessing, feature extraction, and detection stages. Each technique processes information from posters and movie subtitles separately. Hence, it proposes a novel model for movie genre classification that can significantly improve the performance of genre detection systems by making them more accurate and reliable. The innovations in approach will help in better identification, which enhances the performance of the systems associated with the movie industry. The implementation results showed that the proposed method achieved a high accuracy of 92% and an F-measure of 91%, indicating the superiority of the proposed method compared to the comparative methods.