Powering Finetuning in Few-Shot Learning: Domain-Agnostic Bias Reduction with Selected Sampling
This work addresses the challenge of improving few-shot learning accuracy for practical applications by reducing biases in feature distributions, though it is incremental as it builds on existing finetuning methods.
The paper tackles the problem of refining novel-class features in few-shot learning by reducing class-agnostic and class-specific biases during finetuning, achieving state-of-the-art results on Meta-Dataset with consistent performance boosts across ten datasets.
In recent works, utilizing a deep network trained on meta-training set serves as a strong baseline in few-shot learning. In this paper, we move forward to refine novel-class features by finetuning a trained deep network. Finetuning is designed to focus on reducing biases in novel-class feature distributions, which we define as two aspects: class-agnostic and class-specific biases. Class-agnostic bias is defined as the distribution shifting introduced by domain difference, which we propose Distribution Calibration Module(DCM) to reduce. DCM owes good property of eliminating domain difference and fast feature adaptation during optimization. Class-specific bias is defined as the biased estimation using a few samples in novel classes, which we propose Selected Sampling(SS) to reduce. Without inferring the actual class distribution, SS is designed by running sampling using proposal distributions around support-set samples. By powering finetuning with DCM and SS, we achieve state-of-the-art results on Meta-Dataset with consistent performance boosts over ten datasets from different domains. We believe our simple yet effective method demonstrates its possibility to be applied on practical few-shot applications.