An Evaluation of Self-Supervised Pre-Training for Skin-Lesion Analysis
Authors:
- Levy ChavesLCLevy Chaves
- Alceu BissotoABAlceu Bissoto
- Eduardo ValleEVEduardo Valle
- Sandra AvilaSandra Avila
Preprint
Abstract
Self-supervised pre-training appears as an advantageous alternative to supervised pre-trained for transfer learning. By synthesizing annotations on pretext tasks, self-supervision allows to pre-train models on large amounts of pseudo-labels before fine-tuning them on the target task. In this work, we assess self-supervision for the diagnosis of skin lesions, comparing three self-supervised pipelines to a challenging supervised baseline, on five test datasets comprising in- and out-of-distribution samples. Our results show that self-supervision is competitive both in improving accuracies and in reducing the variability of outcomes. Self-supervision proves particularly useful for low training data scenarios ($<1\,500$ and $<150$ samples), where its ability to stabilize the outcomes is essential to provide sound results.
Keywords
All Related Versions
This publication has 12 references indexed in Scilit:
- Momentum contrastive learning for few-shot COVID-19 diagnosis from chest CT imagesPattern Recognition, 2021
- Unlabeled skin lesion classification by self-supervised topology clustering networkBiomedical Signal Processing and Control, 2021
- A patient-centric dataset of images and metadata for identifying melanomas using clinical contextScientific Data, 2021
- PAD-UFES-20: A skin lesion dataset composed of patient data and clinical images collected from smartphonesData in Brief, 2020
- Comparing to Learn: Surpassing ImageNet Pretraining on Radiographs by Comparing Image Representations2020
- Data, depth, and design: Learning reliable models for skin lesion analysisNeurocomputing, 2020
- Self-Supervised Learning for Cardiac MR Image Segmentation by Anatomical Position Prediction2019
- Self-supervised learning for medical image analysis using image context restorationMedical Image Analysis, 2019
- Self-supervised Learning for Dense Depth Estimation in Monocular Endoscopy2018
- Retinal Image Understanding Emerges from Self-Supervised Multimodal Reconstruction2018