Fig. 5
From: Multi-sample \(\zeta \)-mixup: richer, more realistic synthetic samples from a p-series interpolant
Visualizing the results obtained using mixup  (b) and \(\zeta \)-mixup  (c, d, e) on images (a) from the ISIC 2018 dataset, with three values of \(\gamma \ (2.4, 2.8, 4.0)\) used for \(\zeta \)-mixup. Similar to Fig. 4, mixup  synthesizes unrealistic images with ghosting (selected images highlighted in blue in b), with multiple lesions overlapping, with artifacts (hair) overlapping the lesion, or with unrealistic anatomical arrangements (lesion, hair overflowing outside the body). And as before, for all values of \(\gamma \), \(\zeta \)-mixup  produces more realistic images