A Unified Framework for Summarization of Nollywood Movie Sequences and Audio Sound Data
Ogar Tumenayu Ofut, Francis Bakpo Sunday, Hyacinth Agozie Eneh, Umoh Enoima Essien · International Journal of Computer Applications · 2022
In recent times, the robustness of movie sources for information and consumptions has attained an unprecedented height.If a successful extraction framework of valuable and interesting information from a movie without the viewer going through the entire movie to understand the storyline is obtainable that is referred to as automatic movie summarization.This act of automatic summarization can foster the management of the growing volume of movie data.This paper shall border on the design of a unified framework for summarizing both movie sequences and its corresponding audio sound data for an automatic summarization task of Nollywood movies.The framework is a Recurrent Neural Network (RNN) model with two layers that handles the selection of key movie sequences, while the audio sound processing and concatenating selected keyframe is handled by Fast Forward Moving Picture Expert Group (FFMPEG).The proposedframework produced good results on the two benchmark datasets, which shows that the experimentation of the new model is properly executed.