为避免人为因素对人脸面部图像皮肤纹理特征提取产生的影响，用卷积神经网络算法对人脸图像修饰进行检测.传统的图像分类方法需要进行复杂的人工特征提取，而卷积神经网络可以自动学习并直接从图像中获取特征，解决了传统模式识别方法特征提取难的问题，具有更高的识别率和更广泛的实用性.在传统卷积神经网络模型中，调整卷积核大小、减少参数、改变卷积层滤波器数量、调整卷积层和池化层的交替方式、使用dropout来提高模型泛化能力以形成适用于人脸修饰检测的新的网络模型.实验结果表明，在引入的数据集上，新的网络模型对人脸图像的修饰检测有较强的鲁棒性，达到了较高的识别率.
In order to avoid the influence of human factors on skin texture feature extraction of facial images, this paper proposes to detect facial image modification by using convolutional neural network (CNN) algorithm. To the best of our knowledge, this is the first report to use CNN in the detection of human face tampering. Compared with the traditional image classification methods which need complex artificial feature extraction, CNN can learn automatically, acquire features directly from the image, and reduce the difficulty of extracting features in traditional pattern recognition methods, accordingly, gaining a higher recognition rate and wider practicality at the same time. On the basis of the traditional convolutional neural network model, the proposed method builds a new network model for human face tampering detection by adjusting the size of the convolution kernel, reducing the parameters, changing the number of convolutional layer filters, adjusting the alternate mode of the convolutional layer and the pooling layer, and using dropout to improve the generalization ability of the model. Experimental results show that the new network model performs with a high recognition rate and strong robustness in the tamper detection of facial images.
[1] Kee E, Farid H. A perceptual metric for photo retouching[J]. Proceedings of the National Academy of Sciences, 2011, 108(50):19907-19912.
[2] Kee E, O'brien J F, Farid H. Exposing photo manipulation from shading and shadows[J]. ACM Transaction on Graphics, 2014, 33(5):1-21.
[3] Myers P N, Biocca F A. The elastic body image:the effect of television advertising and programming on body image distortions in young women[J]. Journal of Communication, 1992, 42(3):108-133.
[4] Agliata D, Tantleff-Dunn S. The impact of media exposure on males' body image[J]. Journal of Social and Clinical Psychology, 2004, 23(1):7-22.
[5] Bharati A, Singh R, Vatsa M, et al. Detecting facial retouching using supervised deep learning[J]. IEEE Transactions on Information Forensics and Security, 2016, 11(9):1903-1913.
[6] Nampoothiri V P, Sugitha N. Digital image forgery-a threaten to digital forensics[C]//IEEE International Conference on Circuit, Power and Computing Technologies, 2016:1-6.
[7] Peng F, Wang X. Digital image forgery forensics by using blur estimation and abnormal hue detection[C]//IEEE Photonics and Optoelectronic, 2010:1-4.
[8] Du Z, Li X, Guo Y. Exposing blur kernel form retouch image[C]//IEEE Computer Society International Conference on Computer-Aided Design and Computer Graphics, 2013:407-408.
[9] H S, P S, J K. Retouching detection and steganalysis[J]. International Journal of Engineering Innovations and Research, 2013, 2(6):487-490.
[10] Batool N, Chellappa R. Detection and inpainting of facial wrinkles using texture orientation fields and markov random field modeling[J]. IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society, 2014, 23(9):3773-3788.
[11] Liu X, Xie L, Zhong B, et al. Automatic facial flaw detection and retouching via discriminative structure tensor[J]. IET Image Processing, 2017, 11(11):1068-1076.
[12] Lecun Y, Bengio Y. Convolutional networks for images, speech, and time series[M]. America:MIT press, 1995.
[13] Krizhevsky A, Sutskever Ii, Hinton G. Image net classification with deep convolution neural networks[C]//Advances in Neural Information Processing Systems, 2012:1097-1105.
[14] Zhon F Y, Jin L P, D J. A review of convolutional neural networks[J]. Chinese Journal of Computers, 2017, 40(6):1229-1251.
[15] Yoo H J. Deep convolution neural networks in computer vision[J]. IEEE Transactions on Smart Processing and Computing, 2015, 4(1):35-43.
[16] Jia Y, Shelhamer E, Donahue J, et al. Caffe:convolutional architecture for fast feature embedding[C]//Proceedings of the 22nd ACM International Conference on Multimedia, 2014:675-678.
