ManTra-Net: Manipulation Tracing Network for Detection and Localization of Image Forgeries With Anomalous Features

To fight against real-life image forgery, which commonly involves different types and combined manipulations, we propose a unified deep neural architecture called ManTra-Net. Unlike many existing solutions, ManTra-Net is an end-to-end network that performs both detection and localization without extra preprocessing and postprocessing. \manifold is a fully convolutional network and handles images of arbitrary sizes and many known forgery types such splicing, copy-move, removal, enhancement, and even unknown types. This paper has three salient contributions. We design a simple yet effective self-supervised learning task to learn robust image manipulation traces from classifying 385 image manipulation types. Further, we formulate the forgery localization problem as a local anomaly detection problem, design a Z-score feature to capture local anomaly, and propose a novel long short-term memory solution to assess local anomalies. Finally, we carefully conduct ablation experiments to systematically optimize the proposed network design. Our extensive experimental results demonstrate the generalizability, robustness and superiority of ManTra-Net, not only in single types of manipulations/forgeries, but also in their complicated combinations.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Benchmark
Image Manipulation Localization Casia V1+ ManTraNet Average Pixel F1(Fixed threshold) .180 # 7
Image Manipulation Detection Casia V1+ ManTraNet AUC .644 # 7
Balanced Accuracy .500 # 6
Image Manipulation Detection CocoGlide ManTraNet AUC .778 # 1
Balanced Accuracy .500 # 5
Image Manipulation Localization CocoGlide ManTraNet Average Pixel F1(Fixed threshold) .516 # 4
Image Manipulation Localization Columbia ManTraNet Average Pixel F1(Fixed threshold) .508 # 8
Image Manipulation Detection Columbia ManTraNet AUC .810 # 6
Balanced Accuracy .500 # 7
Image Manipulation Localization COVERAGE ManTraNet Average Pixel F1(Fixed threshold) .317 # 7
Image Manipulation Detection COVERAGE ManTraNet AUC .760 # 4
Balanced Accuracy .500 # 6
Image Manipulation Localization DSO-1 ManTraNet Average Pixel F1(Fixed threshold) .412 # 5
Image Manipulation Detection DSO-1 ManTraNet AUC .874 # 4
Balanced Accuracy .500 # 5

Methods


No methods listed for this paper. Add relevant methods here