Saturation Augmentation

Randomly adjust color vibrancy in images to make your model resilient to saturation differences.

Randomly adjust the vibrancy of the colors in the images.

• Percent: The percentage by which color saturation is randomly increased or decreased. Positive values make colors more vivid, negative values make them more muted.

Example

Original Image

Saturation Augmentation of +30%

Saturation Agumentation of -30%

Learn More

• Why Add Noise to Images for Machine Learning