目的 恶意的图像拼接篡改给名誉、法律、政治等带来一系列的挑战，而现有图像拼接检测算法通常采用参数固定的高通滤波器提取滤波特征进行预处理，这没有考虑图像之间的差异。方法 因此，本文设计自适应残差模块(Adaptive Residuals Module, ARM)凸显拼接篡改痕迹，将卷积运算后的残差多次拼接，且每次拼接后再利用注意力机制实现通道间的非线性交互。然后，使用通道注意力(Squeeze and Excitation, SE)模块以减少由ARM提取残差特征产生的通道之间信息冗余，并以在图像分类领域获得卓越性能的EfficientNet为骨干网络，提出一种新的图像拼接检测算法。结果 实验表明，所提算法在CASIA I，CASIA II，COLUMBIA COLOR，NIST16和FaceForensic++五个公开数据集上分别取得98.95%，98.88%，100%，100%，88.20%的检测准确率，获得比现有算法更高的准确率。提出的ARM将骨干网络EfficientNet在CASIA II 数据集的准确率提高3.94%以上。结论 提出的基于自适应残差的图像拼接检测算法充分考虑图像之间的差异，凸显篡改区域与未篡改区域之间的区别，并获得更好的拼接检测结果。
Adaptive Residual Algorithm for Image Splicing Detection
ZHANG LING, MU WENPENG1, CHEN BEIJING2(1.单位,所在省市 邮编,国家;2.Nanjing University of Information Science and Technology)
Objective In recent years, digital media have become central to the exchange of information in our daily lives. With the rapid development of image editing tools and deep learning techniques, it is easy to tamper with transmitted images. Image splicing is one of the most common types of image tampering. Malicious image splicing challenges reputation, law, and politics. Therefore, various approaches have been proposed for detecting image splicing forgeries. Deep learning has also been successfully applied in image splicing detection. However, the existing deep learning-based works usually preprocess the input images by extracting the filtered features by the high-pass filters with fixed parameters, which does not consider the differences between images. Methods Therefore, a new image splicing detection algorithm is proposed in this paper. Firstly, an adaptive residual module (ARM) is designed to highlight the splicing traces. In the ARM, the residual after the convolution operation is serialized several times and the attention mechanism is used to realize the nonlinear interaction between channels after each connection. Unlike ordinary filters with fixed parameters, the ARM module entirely relies on the feature reuse and attention mechanism of residuals to retain and enlarge the details of the splicing. Then, a squeeze and excitation (SE) module is used to reduce the inter-channel information redundancy generated by ARM residual feature extraction. The SE module uses an average adaptive pool to generate channel statistics information on global space and the gating mechanism of the Sigmoid activation function to learn channel weights from channel dependencies. Finally, a new image splicing detection algorithm is proposed by combining with the proposed ARM and the backbone network EffcientNet, a model with excellent performance in image classification. Results Experimental results show that the proposed algorithm achieves 98.95%, 98.88%, 100%, 100%, and 88.20% detection accuracies on CASIA I, CASIA II, COLUMBIA COLOR, NIST16, and FaceForensic++, respectively, and get higher accuracy than the existing algorithms. Besides, the proposed ARM algorithm improves the accuracy of backbone network by 3.94% on the CASIA II dataset. Regarding the computational time, on the CASIA II dataset, the training time per batch of the proposed algorithm is 71.75s, and the test time for a single image is 0.011s, which is less than the existing algorithms. Besides, the size of the parameters of ARM is 0.0036MB, which is about 2? of the parameters size of the backbone network EfficientNet, and the FLOPs are about 0.037G. Conclusion This paper proposes an image splicing detection algorithm based on ARM, and the proposed algorithm performs well on five public datasets. The designed ARM is a plug-and-play lightweight, adaptive feature extraction module. It can be migrated on other models, such as Xception and ResNet, etc.