MetAdapt: Meta-Learned Task-Adaptive Architecture for Few-Shot Classification

TL;DR

This paper introduces MetAdapt, a meta-learned, task-adaptive neural architecture search method for few-shot classification, achieving state-of-the-art results by optimizing architectures and adapting them at test time.

Contribution

It combines meta-learning with neural architecture search to optimize and adapt network architectures specifically for few-shot learning tasks.

Findings

Achieves state-of-the-art results on miniImageNet and FC100 benchmarks.

Introduces MetAdapt Controller modules for task-specific architecture adaptation.

Demonstrates the effectiveness of architecture optimization in few-shot learning.

Abstract

Few-Shot Learning (FSL) is a topic of rapidly growing interest. Typically, in FSL a model is trained on a dataset consisting of many small tasks (meta-tasks) and learns to adapt to novel tasks that it will encounter during test time. This is also referred to as meta-learning. Another topic closely related to meta-learning with a lot of interest in the community is Neural Architecture Search (NAS), automatically finding optimal architecture instead of engineering it manually. In this work, we combine these two aspects of meta-learning. So far, meta-learning FSL methods have focused on optimizing parameters of pre-defined network architectures, in order to make them easily adaptable to novel tasks. Moreover, it was observed that, in general, larger architectures perform better than smaller ones up to a certain saturation point (where they start to degrade due to over-fitting). However, little attention has been given to explicitly optimizing the architectures for FSL, nor to an adaptation of the architecture at test time to particular novel tasks. In this work, we propose to employ tools inspired by the Differentiable Neural Architecture Search (D-NAS) literature in order to optimize the architecture for FSL without over-fitting. Additionally, to make the architecture task adaptive, we propose the concept of `MetAdapt Controller' modules. These modules are added to the model and are meta-trained to predict the optimal network connections for a given novel task. Using the proposed approach we observe state-of-the-art results on two popular few-shot benchmarks: miniImageNet and FC100.