Hands-On One-shot Learning with Python

Image segmentation in few-shot domains

The research work CANet: Class-agnostic segmentation networks with iterative refinement and attentive few-shot learning (https://arxiv.org/abs/1903.02351) is proof of potential growth in the medical imaging industry. In this paper, the authors proposed a two-level framework for semantic segmentation: a dense comparison module (DCM) and an iterative optimization module (IOM). DCM does dense feature comparison among training-set examples and test-set examples by extracting features using common ResNet architecture, whereas IOM refines results over iteration through a residual block+CNN and an atrous spatial pyramid pooling (ASPP) module.

Similarly, PANet: Few-shot image semantic segmentation with prototype ...