A Universal Representation Transformer Layer for Few-Shot Image Classification

by   Lu Liu, et al.
Apple, Inc.

Few-shot classification aims to recognize unseen classes when presented with only a small number of samples. We consider the problem of multi-domain few-shot image classification, where unseen classes and examples come from diverse data sources. This problem has seen growing interest and has inspired the development of benchmarks such as Meta-Dataset. A key challenge in this multi-domain setting is to effectively integrate the feature representations from the diverse set of training domains. Here, we propose a Universal Representation Transformer (URT) layer, that meta-learns to leverage universal features for few-shot classification by dynamically re-weighting and composing the most appropriate domain-specific representations. In experiments, we show that URT sets a new state-of-the-art result on Meta-Dataset. Specifically, it outperforms the best previous model on three data sources or performs the same in others. We analyze variants of URT and present a visualization of the attention score heatmaps that sheds light on how the model performs cross-domain generalization. Our code is available at https://github.com/liulu112601/URT.


page 1

page 2

page 3

page 4


Universal Representation Learning from Multiple Domains for Few-shot Classification

In this paper, we look at the problem of few-shot classification that ai...

Improving Task Adaptation for Cross-domain Few-shot Learning

In this paper, we look at the problem of cross-domain few-shot classific...

Combining Domain-Specific Meta-Learners in the Parameter Space for Cross-Domain Few-Shot Classification

The goal of few-shot classification is to learn a model that can classif...

Learning a Universal Template for Few-shot Dataset Generalization

Few-shot dataset generalization is a challenging variant of the well-stu...

Universal Representations: A Unified Look at Multiple Task and Domain Learning

We propose a unified look at jointly learning multiple vision tasks and ...

Tree Structure-Aware Few-Shot Image Classification via Hierarchical Aggregation

In this paper, we mainly focus on the problem of how to learn additional...

Enhancing Few-shot Image Classification with Cosine Transformer

This paper addresses the few-shot image classification problem. One nota...

Please sign up or login with your details

Forgot password? Click here to reset