Few-shot classification

Metric learning methods are widely used in few-shot learning due to their simplicity and effectiveness. Most existing methods directly …

Few-shot classification aims to identify novel categories using only a few labeled samples. Generally, the metric-based few-shot …