FSL-Rectifier: Rectify Outliers in Few-Shot Learning via Test-Time Augmentation
This addresses generalization challenges in few-shot learning for computer vision applications, but it is incremental as it builds on existing augmentation methods.
The paper tackles the problem of outlier samples in few-shot learning by generating additional test-class samples via a generative image combiner and averaging features to reduce bias, resulting in a test accuracy improvement of around 10% (e.g., from 46.86% to 53.28%) for trained models.
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. So far, plenty of algorithms involve training data augmentation to improve the generalization capability of FSL models, but outlier queries or support images during inference can still pose great generalization challenges. In this work, to reduce the bias caused by the outlier samples, we generate additional test-class samples by combining original samples with suitable train-class samples via a generative image combiner. Then, we obtain averaged features via an augmentor, which leads to more typical representations through the averaging. We experimentally and theoretically demonstrate the effectiveness of our method, obtaining a test accuracy improvement proportion of around 10\% (e.g., from 46.86\% to 53.28\%) for trained FSL models. Importantly, given a pretrained image combiner, our method is training-free for off-the-shelf FSL models, whose performance can be improved without extra datasets nor further training of the models themselves. Codes are available at https://github.com/WendyBaiYunwei/FSL-Rectifier-Pub.