AI engineering glossary · Computer vision
What is cutout?
Cutout (also called Random erasing): An augmentation that blanks out a random square patch of the input image. It forces the model not to depend on one region and improves robustness to occlusion. It acts as a cheap regulariser during training.
Cutout: an example
A 32 x 32 patch of a 224 x 224 photo is set to zero so the model must recognise the object from the remaining parts.
Related terms
More Computer vision terms
- Data augmentation for images
- DeepLab
- Depthwise separable convolution
- Dice coefficient
- Dilated convolution
- Document layout analysis
- Edge detection
- EfficientNet
Where you learn it
Cutout is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.