MDFM: Multi-Decision Fusing Model for Few-Shot Learning

TL;DR

This paper introduces MDFM, a multi-decision fusion approach that enhances few-shot learning by combining decisions from multiple feature extraction models, improving robustness and accuracy across various datasets.

Contribution

The paper proposes a novel, simple, non-parametric multi-decision fusion method for few-shot learning that can be applied to existing models and extends to supervised and semi-supervised settings.

Findings

Achieves 3.4%-7.3% improvements over state-of-the-art methods.

Effective in both supervised and semi-supervised few-shot learning.

Validated on five benchmark datasets.

Abstract

In recent years, researchers pay growing attention to the few-shot learning (FSL) task to address the data-scarce problem. A standard FSL framework is composed of two components: i) Pre-train. Employ the base data to generate a CNN-based feature extraction model (FEM). ii) Meta-test. Apply the trained FEM to the novel data (category is different from base data) to acquire the feature embeddings and recognize them. Although researchers have made remarkable breakthroughs in FSL, there still exists a fundamental problem. Since the trained FEM with base data usually cannot adapt to the novel class flawlessly, the novel data's feature may lead to the distribution shift problem. To address this challenge, we hypothesize that even if most of the decisions based on different FEMs are viewed as weak decisions, which are not available for all classes, they still perform decently in some specific categories. Inspired by this assumption, we propose a novel method Multi-Decision Fusing Model (MDFM), which comprehensively considers the decisions based on multiple FEMs to enhance the efficacy and robustness of the model. MDFM is a simple, flexible, non-parametric method that can directly apply to the existing FEMs. Besides, we extend the proposed MDFM to two FSL settings (i.e., supervised and semi-supervised settings). We evaluate the proposed method on five benchmark datasets and achieve significant improvements of 3.4%-7.3% compared with state-of-the-arts.