Learning Class-level Prototypes for Few-shot Learning

2021/08/25 by Minglei Yuan, Yuan, Minglei, Wenhai Wang +9
Computer Science · #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning and Algorithms

paper · pdf · doi:10.48550/arxiv.2108.11072

openalex publication_date 2021/08/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Few-shot learning aims to recognize new categories using very few labeled samples. Although few-shot learning has witnessed promising development in recent years, most existing methods adopt an average operation to calculate prototypes, thus limited by the outlier samples. In this work, we propose a simple yet effective framework for few-shot classification, which can learn to generate preferable prototypes from few support data, with the help of an episodic prototype generator module. The generated prototype is meant to be close to a certain \textit\targetproto and is less influenced by outlier samples. Extensive experiments demonstrate the effectiveness of this module, and our approach gets a significant raise over baseline models, and get a competitive result compared to previous methods on miniImageNet, tieredImageNet, and cross-domain (miniImageNet → CUB-200-2011) datasets.

Related

• PrototypeFormer: Learning to Explore Prototype Relationships for Few-shot Image Classification
• A Closer Look at Few-shot Classification
• Extended Few-Shot Learning: Exploiting Existing Resources for Novel Tasks
• Few and Fewer: Learning Better from Few Examples Using Fewer Base Classes
• Few-shot Learning via Dependency Maximization and Instance Discriminant Analysis