| Title | Recaptured Image Forensics Based on Generalized Central Difference Convolution Network |
| Publication Type | Conference Paper |
| Year of Publication | 2022 |
| Authors | Liu, Zhiqin, Zhu, Nan, Wang, Kun |
| Conference Name | 2022 IEEE 2nd International Conference on Software Engineering and Artificial Intelligence (SEAI) |
| Date Published | jun |
| Keywords | central difference convolution, Conferences, convolution, Deep Learning, feature extraction, Fuses, Human Behavior, human factors, Image databases, Image forensics, information forensics, Liquid crystal displays, Metrics, pubcrawl, recaptured image detection, resilience, Resiliency, Scalability |
| Abstract | With large advancements in image display technology, recapturing high-quality images from high-fidelity LCD screens becomes much easier. Such recaptured images can be used to hide image tampering traces and fool some intelligent identification systems. In order to prevent such a security loophole, we propose a recaptured image detection approach based on generalized central difference convolution (GCDC) network. Specifically, by using GCDC instead of vanilla convolution, more detailed features can be extracted from both intensity and gradient information from an image. Meanwhile, we concatenate the feature maps from multiple GCDC modules to fuse low-, mid-, and high-level features for higher performance. Extensive experiments on three public recaptured image databases demonstrate the superior of our proposed method when compared with the state-of-the-art approaches. |
| DOI | 10.1109/SEAI55746.2022.9832331 |
| Citation Key | liu_recaptured_2022 |
Recaptured Image Forensics Based on Generalized Central Difference Convolution Network