FeLMi : Few shot Learning with hard Mixup

Abstract

Learning from a few examples is a challenging computer vision task. Traditionally, meta-learning-based methods have shown promise towards solving this problem. Recent approaches show benefits by learning a feature extractor on the abundant base examples and transferring these to the fewer novel examples. However, the finetuning stage is often prone to overfitting due to the small size of the novel dataset. To this end, we propose Few shot Learning with hard Mixup (FeLMi) using manifold mixup to synthetically generate samples that helps in mitigating the data scarcity issue. Different from a naïve mixup, our approach selects the hard mixup samples using an uncertainty-based criteria. To the best of our knowledge, we are the first to use hard-mixup for the few-shot learning problem. Our approach allows better use of the pseudo-labeled base examples through base-novel mixup and entropy-based filtering. We evaluate our approach on several common few-shot benchmarks - FC-100, CIFAR-FS, miniImageNet and tieredImageNet and obtain improvements in both 1-shot and 5-shot settings. Additionally, we experimented on the cross-domain few-shot setting (miniImageNet → CUB) and obtain significant improvements.