Pengenalan Trailer Film Melalui Machine Learning Dan Deep Learning Dengan Normalisasi Sequence

Yohanes, Anom Pratitis (2020) Pengenalan Trailer Film Melalui Machine Learning Dan Deep Learning Dengan Normalisasi Sequence. Undergraduate Thesis thesis, Institut Teknologi Telkom Purwokerto.

Preview	Text ABSTRACT.pdf - Accepted Version Download (6kB) | Preview
Preview	Text ABSTRAK.pdf - Accepted Version Download (7kB) | Preview
	Text COVER.pdf - Accepted Version Download (2MB)
Preview	Text BAB 1.pdf - Accepted Version Download (132kB) | Preview
Preview	Text BAB 2.pdf - Accepted Version Download (580kB) | Preview
Preview	Text BAB 3.pdf - Accepted Version Download (499kB) | Preview
	Text BAB 4.pdf - Accepted Version Restricted to Registered users only Download (292kB)
Preview	Text BAB 5.pdf - Accepted Version Download (8kB) | Preview
Preview	Text DAFTAR PUSTAKA.pdf - Accepted Version Download (148kB) | Preview
	Text LAMPIRAN.pdf - Accepted Version Restricted to Registered users only Download (278kB)

Abstract

Watching a film on Youtube or in the cinema has become a natural thing for children, adolescents, adults and parents. Some adults invite their children to watch movies in theaters regardless of the category. Watching action, horror and romance films can each teach children to perform scenes of violence, murder and kissing. Based on these problems, a tool is needed to detect movie trailers using machine learning and deep learning methods. In this study, using 18 videos of people standing still, walking and running. The video will be split into 2,719 dataset images. Furthermore, the image dataset is trained using one of the deep learning methods, namely the Convolutional Neural Network algorithm to produce a dictionary that will be used to detect movie trailers. By using 51 action film trailers, horror and romance were converted into 4,2719 image datasets and trained using the Convolutional Neural Network algorithm to produce sequence values. Then the sequence value is tested using several Machine Learning methods, namely k-nearest neighbor, support vector machine, extra tree classifier, random forest classifier, adaboost classifier, and gradient boosting classifier to get an accuracy value as a faster comparison using machine learning methods or deep learning methods. The results of training the accuracy value of the film trailer dataset using machine learning and deep learning methods are 93% and 100%, respectively. The results of the accuracy value test using machine learning and deep learning methods were 46% and 100%, respectively. The application of the film trailer recognition system through machine learning and deep learning can help the public in detecting movie trailers with the right accuracy. Keywords : Augmentation, Deep Learning, Film Trailer Detection, Machine Learning

Actions (login required)

View Item