Abstract Image inpainting networks based on deep learning techniques have been widely used in many important fields. However, most inpainting networks fail to generate desirable repaired images. This may be due to their failure to extract effective features and accurately assign high weights to the undamaged regions. To alleviate these problems, an image inpainting network based on gated convolution and multi‐level attention mechanism (IIN‐GCMAM) is proposed in this paper. This network follows encoder–decoder architecture, consisting of the gated convolution encoder (GC‐encoder) and the multi‐level attention mechanism decoder (MAM‐decoder). The GC‐encoder weighs the extracted features with gated convolutions, which reduces the interference caused by the damaged regions. The multi‐level attention mechanism employed in the MAM‐decoder uses multi‐scale feature maps spatially and channel‐wise to improve the consistency in global structure and the fineness of repaired results. Extensive experiments are conducted on the common datasets, Paris StreetView and CelebA. Experimental results indicate that the proposed IIN‐GCMAM can achieve a good performance on the common evaluation metrics and visual effects. It can achieve 0.0408, 0.720, and 22.27 in MAE, SSIM, and PSNR at the mask ratio of 50%–60%, respectively.

Channel attention has been shown to improve the performance of deep convolutional neural networks efficiently. Channel attention adaptively recalibrates the importance of each channel, determining what to attend to. However, channel attention only encodes inter-channel information but neglects the importance of positional information. Positional information is crucial in determining where to attend to. To address this issue, we propose a novel channel-spatial attention method named Spatial-Decomposition-Aggregation Attention (SDAA) method. First, a high-axis spatial direction is decomposed into multiple low-axis spatial directions. Then, a shared transformation sub-unit establishes attention in each low-axis space direction. Next, all the low-axis attention masks are aggregated into a high-axis attention mask. Finally, the generated high-axis attention mask is fused into the input features, thus enhancing the input features. Essentially, our method is a divide-and-conquer process. Experimental results demonstrate that our SDAA method outperforms the existing channel-spatial attention methods.

Repackaged Android apps are the major source of Android malware, which not only compromise the pecuniary profit of original authors, but also pose threat to security and privacy of mobile users. Although a large number of birthmark based approaches have been proposed for Android repackaging detection, the majority of them heavily rely on the code instruction details, thus suffering from the following two limitations: (1) subject to code/resource obfuscation technologies; (2) fail to large scale repackaging detection. In this paper, we propose a novel behavior based approach for Android repackaging detection to meet scalability and obfuscation-resilience at the same time. As the repackaged app always keeps the basic functionalities of the original one for leveraging its popularity, they usually have similar behaviors. This observation inspires us to design the new behavior based birthmark for Android repackaging detection, namely, API dependency graph. To further improve the detection performance, we also introduce a system dependency summary graph based ADG extraction approach for high efficiency birthmark construction. We implement a prototype system named ACFinder and evaluate our system using 13,917 apps of 22 categories collected from APK-DL. Experiments show that ACFinder can extract behavior birthmark efficiently (average 52.9s per app), and that our behavior birthmark is resilient to complex code obfuscation technologies (average app similarity all are 1.0 for 11 code obfuscation algorithms) and capable to large scale detection (average 0.37s per app pair).