Eu Wern Teh, Terrance DeVries, Graham W. Taylor
We consider the problem of distance metric learning (DML), where the task is to learn an effective similarity measure between images. We revisit ProxyNCA and incorporate several enhancements. We find that low temperature scaling is a performance-critical component and explain why it works. Besides, we also discover that Global Max Pooling works better in general when compared to Global Average Pooling. Additionally, our proposed fast moving proxies also addresses small gradient issue of proxies, and this component synergizes well with low temperature scaling and Global Max Pooling. Our enhanced model, called ProxyNCA++, achieves a 22.9 percentage point average improvement of Recall@1 across four different zero-shot retrieval datasets compared to the original ProxyNCA algorithm. Furthermore, we achieve state-of-the-art results on the CUB200, Cars196, Sop, and InShop datasets, achieving Recall@1 scores of 72.2, 90.1, 81.4, and 90.9, respectively.
| Task | Dataset | Metric | Value | Model |
|---|---|---|---|---|
| Image Retrieval | CARS196 | R@1 | 90.1 | ProxyNCA++ |
| Image Retrieval | SOP | R@1 | 81.4 | ProxyNCA++ |
| Image Retrieval | In-Shop | R@1 | 90.9 | ProxyNCA++ |
| Image Retrieval | CUB-200-2011 | R@1 | 72.2 | ProxyNCA++ |
| Metric Learning | CARS196 | R@1 | 86.5 | ResNet-50 + ProxyNCA++ |
| Metric Learning | CUB-200-2011 | R@1 | 69 | ResNet-50 + ProxyNCA++ |
| Metric Learning | In-Shop | R@1 | 90.9 | ResNet-50 + ProxyNCA++ |
| Metric Learning | Stanford Online Products | R@1 | 80.7 | ResNet-50 + ProxyNCA++ |