Boosting Few-Shot Learning with Disentangled Self-Supervised Learning and Meta-Learning for Medical Image Classification

arXiv:2403.17530v1 Announce Type: new
Abstract: Background and objective: Employing deep learning models in critical domains such as medical imaging poses challenges associated with the limited availability of training data. We present a strategy for improving the performance and generalization capabilities of models trained in low-data regimes. Methods: The proposed method starts with a pre-training phase, where features learned in a self-supervised learning setting are disentangled to improve the robustness of the representations for downstream tasks. We then introduce a meta-fine-tuning …

abstract arxiv availability boosting capabilities challenges classification cs.ai cs.cv data deep learning domains few-shot few-shot learning image imaging improving low medical medical imaging meta meta-learning performance self-supervised learning strategy supervised learning training training data type