Bevan, Peter and Atapour-Abarghouei, Amir (2022) 'Skin Deep Unlearning: Artefact and Instrument Debiasing in the Context of Melanoma Classification.', The 39th International Conference on Machine Learning (ICML 2022) Baltimore, MD, 17-23 July 2022.
Convolutional Neural Networks have demonstrated dermatologist-level performance in the classification of melanoma and other skin lesions, but prediction irregularities due to biases seen within the training data are an issue that should be addressed before widespread deployment is possible. In this work, we robustly remove bias and spurious variation from an automated melanoma classification pipeline using two leading bias ‘unlearning’ techniques. We show that the biases introduced by surgical markings and rulers presented in previous studies can be reasonably mitigated using these bias removal methods. We also demonstrate the generalisation benefits of ‘unlearning’ spurious variation relating to the imaging instrument used to capture lesion images. Our experimental results provide evidence that the effects of each of the aforementioned biases are notably reduced, with different debiasing techniques excelling at different tasks.
|Item Type:||Conference item (Paper)|
|Full text:||Publisher-imposed embargo |
(AM) Accepted Manuscript
File format - PDF (9388Kb)
|Publisher Web site:||https://proceedings.mlr.press/|
|Date accepted:||15 May 2022|
|Date deposited:||10 June 2022|
|Date of first online publication:||2022|
|Date first made open access:||No date available|
Save or Share this output
|Look up in GoogleScholar|