merges two images by averaging their pixel intensities
The new image has the same training image label opposite to MixUp and other approaches where labels are updated according to the proportion of image mixing.
one epoch on ImageNet and 100 epochs on other datasets are completed without SamplePairing before mixed image data is added to the training
Once the SamplePairing images are added to the training set, they run in cycles between 8:2 epochs, 8 with SamplePairing images, 2 without.