Image Manipulation Detection by Multi-View Multi-Scale Supervision

The key challenge of image manipulation detection is how to learn generalizable features that are sensitive to manipulations in novel data, whilst specific to prevent false alarms on authentic images. Current research emphasizes the sensitivity, with the specificity overlooked. In this paper we address both aspects by multi-view feature learning and multi-scale supervision. By exploiting noise distribution and boundary artifact surrounding tampered regions, the former aims to learn semantic-agnostic and thus more generalizable features. The latter allows us to learn from authentic images which are nontrivial to be taken into account by current semantic segmentation network based methods. Our thoughts are realized by a new network which we term MVSS-Net. Extensive experiments on five benchmark sets justify the viability of MVSS-Net for both pixellevel and image-level manipulation detection.

Automated summary

The key challenge of image manipulation detection lies in learning features that are sensitive to manipulations in novel data while preventing false alarms on authentic images. This paper addresses both sensitivity and specificity through multi-view feature learning and multi-scale supervision, resulting in a new network called MVSS-Net that demonstrates reliability in detecting pixel-level and image-level manipulations.