Detecting GAN-generated Images by Orthogonal Training of Multiple CNNs.
Sara MandelliNicolò BonettiniPaolo BestaginiStefano TubaroPublished in: CoRR (2022)
Keyphrases
- image analysis
- image database
- image data
- three dimensional
- input image
- image registration
- image recognition
- image collections
- object recognition
- rigid body
- image classification
- image features
- keypoints
- multiple images
- edge detection
- multiple objects
- learning algorithm
- small number
- ground truth
- image annotation
- test images
- cellular neural networks
- image set
- lighting conditions
- spatial information
- image regions
- segmentation method
- training examples
- segmentation algorithm
- feature points
- supervised learning