FSL-Rectifier: Rectify Outliers in Few-Shot Learning via Test-Time Augmentation

1 code implementation • 28 Feb 2024 • Yunwei Bai, Ying Kiat Tan, Tsuhan Chen

Few-shot-learning (FSL) commonly requires a model to identify images (queries) that belong to classes unseen during training, based on a few labelled samples of the new classes (support set) as reference.

Data Augmentation Few-Shot Learning +1