Few-shot Metric Learning: Online Adaptation of Embedding for Retrieval
This work addresses a domain-specific challenge in computer vision by enabling more adaptable metric learning for retrieval tasks, though it is incremental as it builds on existing meta-learning and metric learning approaches.
The paper tackles the problem of metric learning's poor generalization to unseen classes with domain gaps by introducing few-shot metric learning, where CRML adapts embeddings online using few annotated data, achieving improved image retrieval performance, especially with larger domain gaps.
Metric learning aims to build a distance metric typically by learning an effective embedding function that maps similar objects into nearby points in its embedding space. Despite recent advances in deep metric learning, it remains challenging for the learned metric to generalize to unseen classes with a substantial domain gap. To tackle the issue, we explore a new problem of few-shot metric learning that aims to adapt the embedding function to the target domain with only a few annotated data. We introduce three few-shot metric learning baselines and propose the Channel-Rectifier Meta-Learning (CRML), which effectively adapts the metric space online by adjusting channels of intermediate layers. Experimental analyses on miniImageNet, CUB-200-2011, MPII, as well as a new dataset, miniDeepFashion, demonstrate that our method consistently improves the learned metric by adapting it to target classes and achieves a greater gain in image retrieval when the domain gap from the source classes is larger.