Few-shot Learning via Dependency Maximization and Instance Discriminant Analysis

TL;DR

This paper introduces a semi-supervised few-shot learning method that leverages unlabeled data through dependency maximization and instance discriminant analysis, improving performance on standard benchmarks.

Contribution

It proposes a novel dependency maximization technique and an instance discriminant analysis for pseudo-labeling and selecting unlabeled data in few-shot learning.

Findings

Outperforms state-of-the-art methods on mini-ImageNet, tiered-ImageNet, CUB, and CIFARFS.

Effective use of unlabeled data enhances few-shot learning accuracy.

Iterative pseudo-label refinement improves model stability.

Abstract

We study the few-shot learning (FSL) problem, where a model learns to recognize new objects with extremely few labeled training data per category. Most of previous FSL approaches resort to the meta-learning paradigm, where the model accumulates inductive bias through learning many training tasks so as to solve a new unseen few-shot task. In contrast, we propose a simple approach to exploit unlabeled data accompanying the few-shot task for improving few-shot performance. Firstly, we propose a Dependency Maximization method based on the Hilbert-Schmidt norm of the cross-covariance operator, which maximizes the statistical dependency between the embedded feature of those unlabeled data and their label predictions, together with the supervised loss over the support set. We then use the obtained model to infer the pseudo-labels for those unlabeled data. Furthermore, we propose anInstance Discriminant Analysis to evaluate the credibility of each pseudo-labeled example and select the most faithful ones into an augmented support set to retrain the model as in the first step. We iterate the above process until the pseudo-labels for the unlabeled data becomes stable. Following the standard transductive and semi-supervised FSL setting, our experiments show that the proposed method out-performs previous state-of-the-art methods on four widely used benchmarks, including mini-ImageNet, tiered-ImageNet, CUB, and CIFARFS.