Unsupervised Meta-Learning via Dynamic Head and Heterogeneous Task Construction for Few-Shot Classification

TL;DR

This paper introduces DHM-UHT, a novel unsupervised meta-learning algorithm that constructs heterogeneous tasks dynamically, demonstrating state-of-the-art results in zero-shot and few-shot classification by leveraging unsupervised task construction and meta-learning strategies.

Contribution

The paper proposes DHM-UHT, a dynamic head meta-learning method with unsupervised heterogeneous task construction, advancing unsupervised meta-learning for few-shot classification.

Findings

DHM-UHT achieves state-of-the-art performance on several datasets.

Meta-learning shows robustness to label noise and task heterogeneity.

Unsupervised task construction enhances few-shot learning effectiveness.

Abstract

Meta-learning has been widely used in recent years in areas such as few-shot learning and reinforcement learning. However, the questions of why and when it is better than other algorithms in few-shot classification remain to be explored. In this paper, we perform pre-experiments by adjusting the proportion of label noise and the degree of task heterogeneity in the dataset. We use the metric of Singular Vector Canonical Correlation Analysis to quantify the representation stability of the neural network and thus to compare the behavior of meta-learning and classical learning algorithms. We find that benefiting from the bi-level optimization strategy, the meta-learning algorithm has better robustness to label noise and heterogeneous tasks. Based on the above conclusion, we argue a promising future for meta-learning in the unsupervised area, and thus propose DHM-UHT, a dynamic head meta-learning algorithm with unsupervised heterogeneous task construction. The core idea of DHM-UHT is to use DBSCAN and dynamic head to achieve heterogeneous task construction and meta-learn the whole process of unsupervised heterogeneous task construction. On several unsupervised zero-shot and few-shot datasets, DHM-UHT obtains state-of-the-art performance. The code is released at https://github.com/tuantuange/DHM-UHT.