Universal Representation Learning from Multiple Domains for Few-shot Classification

By Wei-Hong Li, Xialei Liu, Hakan Bilen. 2021

115 citations recorded by the source corpus.

In this paper, we look at the problem of few-shot image classification that aims to learn a classifier for previously unseen classes and domains from few labeled samples. Recent methods use various adaptation strategies for aligning their visual representations to new domains or select the relevant ones from multiple domain-specific feature extractors. In this work, we present URL, which learns a single set of universal visual representations by distilling knowledge of multiple domain-specific networks after co-aligning their features with the help of adapters and centered kernel alignment. We show that the universal representations can be further refined for previously unseen domains by an efficient adaptation step in a similar spirit to distance learning methods. We rigorously evaluate our model in the recent Meta-Dataset benchmark and demonstrate that it significantly outperforms the previous methods while being more efficient.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Infectious Diseases, Representation & Self-Supervised Learning, Meta & Continual Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Infectious Diseases

In Representation & Self-Supervised Learning

In Meta & Continual Learning

Citations in the corpus

A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunities (2022, background): “[91] inspired by [92], [93], proposes to map domainspecific features to the same shared space, thus achieving a domain-irrelevant universal representation.”
Forward Compatible Training for Large-Scale Embedding Retrieval Systems (2021, unknown): “Methods are variously classified under few-shot [25, 45], continual learning [35, 48], and lifelong learning [28, 39].”
Improving Task Adaptation for Cross-domain Few-shot Learning (2021, methodology): “URL [22] learns a single multi-domain network by distilling the knowledge of domain-specific networks.”