With the rapid development of multimedia, violent video can be easily accessed in games, movies, websites, and so on. Identifying violent videos and rating violence extent is of great importance to media filtering and children protection. Many previous studies only address the problems of violence scene detection and violent action recognition, yet violence rating problem is still not solved. In this paper, we present a novel video-level rating prediction method to estimate violence extent automatically. It has two main characteristics: (1) a two-stream network is fine-tuned to construct effective representations of violent videos; (2) a violence rating prediction machine is designed to learn the strength relationship among different videos. Furthermore, we present a novel violent video dataset with a total of 1,930 human-involved violent videos designed for violence rating analysis. Each video is annotated with 6 fine-grained objective attributes, which are considered to be closely related to violence extent. The ground-truth of violence rating is given by pairwise comparison method. The dataset is evaluated in both stability and convergence. Experiment results on this dataset demonstrate the effectiveness of our method compared with the state-of-art classification methods.
Ying JI
Nagoya University
Yu WANG
Ritsumeikan University
Jien KATO
Ritsumeikan University
Kensaku MORI
Nagoya University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Ying JI, Yu WANG, Jien KATO, Kensaku MORI, "Predicting Violence Rating Based on Pairwise Comparison" in IEICE TRANSACTIONS on Information,
vol. E103-D, no. 12, pp. 2578-2589, December 2020, doi: 10.1587/transinf.2020EDP7056.
Abstract: With the rapid development of multimedia, violent video can be easily accessed in games, movies, websites, and so on. Identifying violent videos and rating violence extent is of great importance to media filtering and children protection. Many previous studies only address the problems of violence scene detection and violent action recognition, yet violence rating problem is still not solved. In this paper, we present a novel video-level rating prediction method to estimate violence extent automatically. It has two main characteristics: (1) a two-stream network is fine-tuned to construct effective representations of violent videos; (2) a violence rating prediction machine is designed to learn the strength relationship among different videos. Furthermore, we present a novel violent video dataset with a total of 1,930 human-involved violent videos designed for violence rating analysis. Each video is annotated with 6 fine-grained objective attributes, which are considered to be closely related to violence extent. The ground-truth of violence rating is given by pairwise comparison method. The dataset is evaluated in both stability and convergence. Experiment results on this dataset demonstrate the effectiveness of our method compared with the state-of-art classification methods.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2020EDP7056/_p
Copy
@ARTICLE{e103-d_12_2578,
author={Ying JI, Yu WANG, Jien KATO, Kensaku MORI, },
journal={IEICE TRANSACTIONS on Information},
title={Predicting Violence Rating Based on Pairwise Comparison},
year={2020},
volume={E103-D},
number={12},
pages={2578-2589},
abstract={With the rapid development of multimedia, violent video can be easily accessed in games, movies, websites, and so on. Identifying violent videos and rating violence extent is of great importance to media filtering and children protection. Many previous studies only address the problems of violence scene detection and violent action recognition, yet violence rating problem is still not solved. In this paper, we present a novel video-level rating prediction method to estimate violence extent automatically. It has two main characteristics: (1) a two-stream network is fine-tuned to construct effective representations of violent videos; (2) a violence rating prediction machine is designed to learn the strength relationship among different videos. Furthermore, we present a novel violent video dataset with a total of 1,930 human-involved violent videos designed for violence rating analysis. Each video is annotated with 6 fine-grained objective attributes, which are considered to be closely related to violence extent. The ground-truth of violence rating is given by pairwise comparison method. The dataset is evaluated in both stability and convergence. Experiment results on this dataset demonstrate the effectiveness of our method compared with the state-of-art classification methods.},
keywords={},
doi={10.1587/transinf.2020EDP7056},
ISSN={1745-1361},
month={December},}
Copy
TY - JOUR
TI - Predicting Violence Rating Based on Pairwise Comparison
T2 - IEICE TRANSACTIONS on Information
SP - 2578
EP - 2589
AU - Ying JI
AU - Yu WANG
AU - Jien KATO
AU - Kensaku MORI
PY - 2020
DO - 10.1587/transinf.2020EDP7056
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 12
JA - IEICE TRANSACTIONS on Information
Y1 - December 2020
AB - With the rapid development of multimedia, violent video can be easily accessed in games, movies, websites, and so on. Identifying violent videos and rating violence extent is of great importance to media filtering and children protection. Many previous studies only address the problems of violence scene detection and violent action recognition, yet violence rating problem is still not solved. In this paper, we present a novel video-level rating prediction method to estimate violence extent automatically. It has two main characteristics: (1) a two-stream network is fine-tuned to construct effective representations of violent videos; (2) a violence rating prediction machine is designed to learn the strength relationship among different videos. Furthermore, we present a novel violent video dataset with a total of 1,930 human-involved violent videos designed for violence rating analysis. Each video is annotated with 6 fine-grained objective attributes, which are considered to be closely related to violence extent. The ground-truth of violence rating is given by pairwise comparison method. The dataset is evaluated in both stability and convergence. Experiment results on this dataset demonstrate the effectiveness of our method compared with the state-of-art classification methods.
ER -