A Novel Violent Video Detection Scheme Based on Modified 3D Convolutional Neural Networks

Abstract

Violent video constitutes a threat to public security, and effective detection algorithms are in urgent need. In order to improve the detection accuracy of 3D convolutional neural networks (3D ConvNet), a novel violent video detection scheme based on the modified 3D ConvNet is proposed. In this paper, the preprocessing method of data is improved, and a new sampling method by using the key frame as dividing nodes is designed. Then, a random sampling method is adapted to produce the input frame sequence. With experimental evaluations on the crowd violence dataset, the results demonstrate the effectiveness of the proposed new sampling method. For three public violent detection datasets: hockey fight, movies, and crowd violence, individualized strategies are implemented to suit the varied clip length. For the short clips, the 3D ConvNet is constructed by using the uniform sampling method. For the longer clips, the new frame sampling strategy is adopted. The proposed scheme obtains competitive results: 99.62% on hockey fight, 99.97% on movies, and 94.3% on crowd violence. The experimental results show that our method is simple and effective.

Keywords

• Violent video detection
• 3D ConvNet
• key frame extraction