Record Details

Detecting violent scenes in movies using Gated Recurrent Units and Discrete Wavelet Transform

Register: Jurnal Ilmiah Teknologi Sistem Informasi

View Archive Info
 
 
Field Value
 
Title Detecting violent scenes in movies using Gated Recurrent Units and Discrete Wavelet Transform
 
Creator Imah, Elly Matul
Laksono, Ivan Kurnia
Karisma, Karisma
Wintarti, Atik
 
Subject deep learning; gate recurrent unit; violence detection; video processing; wavelet
 
Description The easiness of accessing video on various platforms can negatively impact if not done wisely, especially for children. Parental supervision is needed so that movies platforms avoid inappropriate displays such as violence. Violent scenes in movies can trigger children to commit acts of violence, which is not desired. Unfortunately, it is not easy to supervise them fully. This study proposed a method for automatic detection of violent scenes in movies. Automatic violence detection assists the parents and censorship institutions in detecting violence easily. This study uses Gated Recurrent Units (GRU) algorithm and wavelet as feature extraction to detect violent scenes. This paper shows comparative studies on the variation of the mother wavelet. The experimental results show that GRU is robust and deliver the best performance accuracy of 0.96 while combining with mother wavelet Symlet and Coiflets8. The combination of GRU with wavelet Coiflets8 shows better results than the predecessor.
 
Publisher Information Systems - Universitas Pesantren Tinggi Darul Ulum
 
Contributor the Ministry of Education, Culture, Research, and Technology of the Republic of Indonesia
 
Date 2022-04-06
 
Type info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
Peer-reviewed Article
 
Format application/pdf
 
Identifier https://journal.unipdu.ac.id/index.php/register/article/view/2541
10.26594/register.v8i2.2541
 
Source Register: Jurnal Ilmiah Teknologi Sistem Informasi; Vol 8, No 2 (2022): In progress (July); 94-103
Register: Jurnal Ilmiah Teknologi Sistem Informasi; Vol 8, No 2 (2022): In progress (July); 94-103
2502-3357
2503-0477
10.26594/register.v8i2
 
Language eng
 
Relation https://journal.unipdu.ac.id/index.php/register/article/view/2541/pdf
 
Rights Copyright (c) 2022 The Author(s)
http://creativecommons.org/licenses/by-nc-sa/4.0