基于深度学习的视频篡改检测方法

Video Tampering Detection Method Based on Deep Learning

随着新一代信息技术的迅猛发展,视频图像处理的多媒体技术应用越发广泛,数字视频的真实性面临巨大挑战。对视频篡改问题进行了研究,提出了一种基于深度学习的检测方法,即将卷积神经网络(CNN)和长短期记忆网络(LSTM)相结合,融入哈希(Hash)算法,利用卷积神经网络提取特征值,输入至长短期记忆网络,通过全连接层Softmax提取时空特征构建了CNNH-LSTM融合网络模型。将该网络模型应用于直播类视频篡改检测分类,实验结果表明,该模型具有良好的特征提取能力、泛化能力,并具有较好的区分性,为视频篡改检测提供了一种新思路。

With the rapid development of the new generation of information technology,the application of multi-media technology for video or image processing is becoming more and more widespread,and the authenticity of digi-tal videos faces huge challenges.In this paper,a video tampering detection algorithm based on deep learning is pro-posed:the Convolutional Neural Network(CNN)is combined with the Long Short-Term Memory(LSTM)network,and the Hash algorithm is integrated,the features extracted by the convolutional neural network are input into the long-term and short-term memory network model,and the spatial and temporal features are extracted through the fully connected layer Softmax to construct CNNH-LSTM in the fusion model.By applying this network model to live video tampering detection and classification,the experimental results show that the model has good feature extraction capability generalization capability,and good distinguishing,providing a new method for video tamper detection.