Figure 2
From: Automated human cell classification in sparse datasets using few-shot learning
A temporal overview of notable few-shot learning techniques proposed within the past 5 years. "Optimization-based" few-shot learning techniques refer to those proposing changes to optimization processes employed by a network. "Metric-based" few-shot learning techniques refer to those proposing a metric from which a similarity score between a set of samples can be obtained from. "Augmented Metric-based" few-shot learning techniques refer to those proposing an augmentation (such as application of a self-supervised or transductive process) to a metric-based few-shot learning technique.
