Exposure augmentation is a color-based data augmentation operation. It randomly adjusts the exposure parameters of images to simulate overexposed and underexposed shooting effects. It enables the model to adapt to images under different lighting conditions and improves the model's inference accuracy in complex lighting scenarios.
