The user interface (UI) of web applications is usually the entry point of web attacks against enterprises and organizations. Finding the UI elements utilized by the intruders is of great importance both for attack interception and web application fixing. Current attack investigation methods targeting web UI either provide rough analysis results or have poor performance in high concurrency scenarios, which leads to heavy manual analysis work. In this paper, we propose TeSec, an accurate attack investigation method for web UI applications. TeSec makes use of two kinds of correlations. The first one, built from annotated audit log partitioned by PID/TID and delimiter-logs, captures the correspondence between audit log entries and web requests. The second one, modeled by an Aho-Corasick automaton built during system testing period, captures the correspondence between requests and the UI elements/events. Leveraging these two correlations, TeSec can accurately and automatically locate the UI elements/events (i.e., the root cause of the alarm) from an alarm, even in high concurrency scenarios. Furthermore, TeSec only needs to be deployed in the server and does not need to collect logs from the client-side browsers. We evaluate TeSec on 12 web applications. The experimental results show that the matching accuracy between UI events/elements and the alarm is above 99.6%. And security analysts only need to check no more than 2 UI elements on average for each individual forensics analysis. The maximum overhead of average response time and audit log space overhead are low (4.3% and 4.6% respectively).

Reconstructing a 3D shape from a single-view image using deep learning has become increasingly popular recently. Most existing methods only focus on reconstructing the 3D shape geometry based on image constraints. The lack of explicit modeling of structure relations among shape parts yields low-quality reconstruction results for structure-rich man-made shapes. In addition, conventional 2D-3D joint embedding architecture for image-based 3D shape reconstruction often omits the specific view information from the given image, which may lead to degraded geometry and structure reconstruction. We address these problems by introducing VGSNet, an encoder-decoder architecture for view-aware joint geometry and structure learning. The key idea is to jointly learn a multimodal feature representation of 2D image, 3D shape geometry and structure so that both geometry and structure details can be reconstructed from a single-view image. To this end, we explicitly represent 3D shape structures as part relations and employ image supervision to guide the geometry and structure reconstruction. Trained with pairs of view-aligned images and 3D shapes, the VGSNet implicitly encodes the view-aware shape information in the latent feature space. Qualitative and quantitative comparisons with the state-of-the-art baseline methods as well as ablation studies demonstrate the effectiveness of the VGSNet for structure-aware single-view 3D shape reconstruction.

This paper presents a view-guided solution for the task of point cloud completion. Unlike most existing methods directly inferring the missing points using shape priors, we address this task by introducing ViPC (view-guided point cloud completion) that takes the missing crucial global structure information from an extra single-view image. By leveraging a framework that sequentially performs effective cross-modality and cross-level fusions, our method achieves significantly superior results over typical existing solutions on a new large-scale dataset we collect for the view-guided point cloud completion task.