hello,
thanks for sharing the code.
i have a doubt regarding how labels are mixed. as per the equation (2) of the paper, both the images and labels are propagated with a mixing proportion. when i looked at the code: train_USCL\data_aug\dataset_wrapper_Ultrasound_Video_Mixup.py
line 77 for example returns the mixed image but original labels. Can you kindly elaborate why this is so?
thanks
hello,
thanks for sharing the code.
i have a doubt regarding how labels are mixed. as per the equation (2) of the paper, both the images and labels are propagated with a mixing proportion. when i looked at the code: train_USCL\data_aug\dataset_wrapper_Ultrasound_Video_Mixup.py
line 77 for example returns the mixed image but original labels. Can you kindly elaborate why this is so?
thanks