MOVIE GENRE CLASSIFICATION BASED ON ITS POSTER USING VGG16 AND DENSENET 169

Abstract

As the size of the entertainment world develops, the interest for approved movie databases is additionally developing. The class of a movie supplies data on its general substance and has various qualities. Hence , it ought to be very much arranged using the qualities of movies, without exclusions in the dataset and their posters. A clever and stylish poster can be the pivot towards the acknowledgment of a movie. Humans have a very unique ability to understand the world around them in an instance. With regards to movies, we can take a glance at a poster, break down its tones, textures, surfaces, faces, appearances, objects, and so on and rapidly figure out what be the issue here or what does the poster represent. In this way, we humans can identify the genre of a movie by just taking a glance at the poster of the movie. We proposed a model that does the exactly same like humans i.e., take the movie poster as its input and predict its genre from the poster of the movie using VGG16 . The dataset that we used is taken from kaggle which contains the poster of the movies released on or before 2017 across 28 different genres. We carried out modified renditions of two standard deep learning models for image classification : VGG-16, and DenseNet-169. In the evaluation, I show that the proposed strategy yields a great performance on various movie posters.