ArtiFact: A Large-Scale Dataset with Artificial and Factual Images for Generalizable and Robust Synthetic Image Detection.
Md Awsafur RahmanBishmoy PaulNajibul Haque SarkerZaber Ibn Abdul HakimShaikh Anowarul FattahPublished in: CoRR (2023)
Keyphrases
- million images
- input image
- image dataset
- image data
- image noise
- image collections
- partial occlusion
- image features
- test images
- image set
- image retrieval
- original images
- geometric distortions
- image pixels
- spot detection
- image regions
- image analysis
- image content
- image classification
- street view
- images with ground truth
- object detectors
- segmentation method
- lighting conditions
- sample images
- image matching
- illumination conditions
- region of interest
- intensity variations
- bounding box
- feature points
- cluttered background
- complex background
- image transformations
- fundamental problem in computer vision
- image structure
- image database
- web images
- orientation estimation
- annotated images
- pixel values
- photo collections
- image annotation
- aerial images
- edge detection
- road signs
- image search
- single image
- keypoints
- accurate localization
- circular hough transform
- object detection
- contrast enhancement
- texture and shape features
- imaging process
- scale invariant features
- pavement distress
- multiscale
- geometric transformations
- feature descriptors
- set of real images
- object recognition
- high resolution
- segmentation algorithm
- face images
- face recognition
- secret image
- color images
- image representation
- gray level
- iris segmentation
- gray value
- accurate registration
- registration errors