DOA-GAN: Dual-Order Attentive Generative Adversarial Network for Image Copy-Move Forgery Detection and Localization.
Ashraful IslamChengjiang LongArslan BasharatAnthony HoogsPublished in: CVPR (2020)
Keyphrases
- copy move forgery
- digital images
- ridgelet transform
- image features
- input image
- image classification
- image analysis
- multiscale
- object detection
- generative model
- detection algorithm
- accurate localization
- image segmentation
- image representation
- wavelet transform
- high resolution
- edge detection
- feature points
- keypoints
- test images
- contrast enhancement