ImagiNet: A Multi-Content Dataset for Generalizable Synthetic Image Detection via Contrastive Learning.
Delyan BoychevRadostin CholakovPublished in: CoRR (2024)
Keyphrases
- object detectors
- multiscale
- image data
- input image
- object detection
- learning process
- multimedia
- template matching
- image content
- complex background
- image dataset
- image classification
- edge detection
- digital images
- active learning
- image analysis
- reinforcement learning
- million images
- segmentation method
- image regions
- hough transform
- image matching
- image set
- partial occlusion
- bounding box
- road signs
- learning algorithm
- weakly labeled
- digital content
- image collections
- detection algorithm
- single image
- feature points
- low level
- feature extraction
- image segmentation