Dual Adversarial Alignment for Realistic Support-Query Shift Few-shot Learning

TL;DR

This paper introduces a new challenging setting called RSQS for support-query shift few-shot learning, where samples undergo multiple distribution shifts, and proposes a dual adversarial alignment framework (DuaL) to address inter-domain bias and intra-domain variance, significantly improving performance.

Contribution

The paper defines the RSQS benchmark for realistic support-query shift few-shot learning and proposes the DuaL framework with adversarial feature alignment to handle complex distribution shifts.

Findings

DuaL outperforms state-of-the-art methods on CIFAR100, mini-ImageNet, and Tiered-ImageNet.

The proposed method effectively reduces inter-domain bias and intra-domain variance.

Benchmark results demonstrate the difficulty of RSQS and the effectiveness of DuaL.

Abstract

Support-query shift few-shot learning aims to classify unseen examples (query set) to labeled data (support set) based on the learned embedding in a low-dimensional space under a distribution shift between the support set and the query set. However, in real-world scenarios the shifts are usually unknown and varied, making it difficult to estimate in advance. Therefore, in this paper, we propose a novel but more difficult challenge, RSQS, focusing on Realistic Support-Query Shift few-shot learning. The key feature of RSQS is that the individual samples in a meta-task are subjected to multiple distribution shifts in each meta-task. In addition, we propose a unified adversarial feature alignment method called DUal adversarial ALignment framework (DuaL) to relieve RSQS from two aspects, i.e., inter-domain bias and intra-domain variance. On the one hand, for the inter-domain bias, we corrupt the original data in advance and use the synthesized perturbed inputs to train the repairer network by minimizing distance in the feature level. On the other hand, for intra-domain variance, we proposed a generator network to synthesize hard, i.e., less similar, examples from the support set in a self-supervised manner and introduce regularized optimal transportation to derive a smooth optimal transportation plan. Lastly, a benchmark of RSQS is built with several state-of-the-art baselines among three datasets (CIFAR100, mini-ImageNet, and Tiered-Imagenet). Experiment results show that DuaL significantly outperforms the state-of-the-art methods in our benchmark.