1 Quantifying uncertainty of deep neural networks in skin lesion classification Pieter Van Molle , Tim Verbelen, Cedric De Boom, Bert Vankeirsbilck, Jonas De Vylder, Bart Diricx, Tom Kimpe, Pieter Simoens, Bart Dhoedt pieter.vanmolle@ugent.be
2 Background Deep learning → SOTA in image classification Can we augment the dermatologist workflow? Skin lesion classification • ISIC Archive • At MICCAI: ISIC Challenge
3 Background Deep learning → SOTA in image classification Can we augment the dermatologist workflow? Skin lesion classification • ISIC Archive • At MICCAI: ISIC Challenge !! Limitations of neural networks • only a point estimate Correctly capturing uncertainty • typically overconfident is indispensable for a single class
4 Background Bayesian modelling → introduces uncertainty in deep learning e.g. MC dropout Can we augment the dermatologist workflow?
5 Background Bayesian modelling → introduces uncertainty in deep learning e.g. MC dropout Can we augment the dermatologist workflow? Contribution Uncertainty metric that leverages MC dropout • based on the overlap between output distributions models doubt • bounded between 0 and 1 interpretable by a dermatologist
6 Quantifying uncertainty Dropout masks Softmax outputs } Output histograms Classi ! er } Calculate the BC uncertainty using the top-2 class histograms ...
7 Results Expectation When the model is confident, it should perform better high accuracy 268 0.8 94 32 43 0.6 Accuracy 0.4 7 57 0.2 0.0 0.0 ]0.0, 0.2] ]0.2, 0.4] ]0.4, 0.6] ]0.6, 0.8] ]0.8, 1.0] low BC uncertainty uncertainty
8 Thank you for your attention Pieter Van Molle , Tim Verbelen, Cedric De Boom, Bert Vankeirsbilck, Jonas De Vylder, Bart Diricx, Tom Kimpe, Pieter Simoens, Bart Dhoedt pieter.vanmolle@ugent.be
Recommend
More recommend