Detecting Video Inter-Frame Forgeries Based on Convolutional Neural Network Models
Xuan Hau Nguyena, Yongjian HUb, Khan Gohar Hayatc, Van Thinh Led, Tu D. Truonge · American Journal of Computer Sciences and Applications · 2020
In the era of information explosion today, videos are easily captured and made viral in a short time and tampering videos have become easier due to editing software. So, the authenticity of videos become very essential. Video inter-frame forgeries are the most common type of video forgery methods. Until now some algorithms have been suggested for detecting inter-frame forgeries based on handicraft features but the accuracy and speed in processing of suggested algorithms are still challenging. In this paper, we are going to put forward a video forgeries detection method for detecting video inter-frame forgeries based on convolutional neural network (CNN) models by retraining the available CNN models trained on ImageNet dataset. The proposed methods based on CNN models which have been retrained to exploit spatial-temporal relationships in a video to robustly detect inter-frame forgeries. And in order to eliminate the errors due by the network, we have proposed a confidence score instead of the raw output score from networks. Through the results of experiments, we have proven that the proposed method has significantly higher efficiency and accuracy than recent methods.