Rearranging Pixels is a Powerful Black-Box Attack for RGB and Infrared Deep Learning Models.
Jary PomponiDaniele DántoniNicolosi AlessandroSimone ScardapanePublished in: IEEE Access (2023)
Keyphrases
- infrared
- black box
- learning models
- learning algorithm
- infrared images
- black boxes
- learning tasks
- infrared imagery
- loss function
- machine learning
- machine learning algorithms
- white box
- semi supervised learning
- target detection and tracking
- classification models
- visible spectrum
- conditional random fields
- learning problems
- time of flight
- color space
- color information
- test cases
- input image
- color images
- focal plane
- microarray
- unsupervised learning
- graphical models
- object oriented
- pairwise
- short wave