Table 1 DAC, TPR, and FPR for the existing CNNs and the proposed CNN with dataset BOSSBase 1.01
From: A convolutional neural network to detect possible hidden data in spatial domain images
Algorithm | S-UNIWARD | WOW | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Payload (bpp) | 0.2 | 0.4 | 0.2 | 0.4 | ||||||||
CNN | DAC | TPR | FPR | DAC | TPR | FPR | DAC | TPR | FPR | DAC | TPR | FPR |
Ye-Net | 60.1 | – | – | 68.7 | – | – | 66.9 | – | – | 76.7 | – | – |
Yedroudj-Net | 63.5 | – | – | 77.4 |  |  | 72.3 | – | – | 81.1 | – | – |
Zhu-Net | 71.4 | – | – | 80.5 | – | – | 76.9 | – | – | 84.1 | – | – |
GBRAS-Net | 73.6 | 87.7 | 29.2 | 87.1 | 91.5 | 17.9 | 80.3 | 89.6 | 19.0 | 89.8 | 91.8 | 14.4 |
Proposed-Net | 79.3 | 90.5 | 21.6 | 93.1 | 95.5 | 12.1 | 90.2 | 92.1 | 16.7 | 94.4 | 92.5 | 12.6 |