A Universal Representation Transformer Layer for Few-Shot Image Classification

Lu Liu, William L. Hamilton, Guodong Long, Jing Jiang, Hugo Larochelle

Abstract: 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 achieves top-performance on the highest number of data sources compared to competing methods. 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.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

One-sentence Summary: code at: https://github.com/liulu112601/URT

Code: [ liulu112601/URT](https://github.com/liulu112601/URT)

Data: [CIFAR-10](https://paperswithcode.com/dataset/cifar-10), [CIFAR-100](https://paperswithcode.com/dataset/cifar-100), [MNIST](https://paperswithcode.com/dataset/mnist), [Meta-Dataset](https://paperswithcode.com/dataset/meta-dataset)

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2006.11702/code)