AI engineering glossary · Computer vision
What is contrastive learning?
Contrastive learning: A training approach that pulls together embeddings of related pairs, such as two augmented views of one image, and pushes apart unrelated ones. It learns useful visual features without labels.
Contrastive learning: an example
Two random crops of the same photo are treated as a positive pair and crops of other photos as negatives.
Related terms
More Computer vision terms
- Convolution
- Convolutional neural network
- Cross-correlation
- CutMix
- Cutout
- Data augmentation for images
- DeepLab
- Depthwise separable convolution
Where you learn it
Contrastive learning is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.