Visualizing How Embeddings Generalize
This work addresses the challenge of generalization in metric learning for researchers, but it is incremental as it primarily provides visualization tools without introducing a new method.
The paper tackles the problem of understanding how deep metric learning embeddings generalize to new classes by introducing visualization tools that illustrate generalization behavior beyond validation accuracy, and it finds that the best performance consistently comes from strategies focusing on a few well-selected triplets.
Deep metric learning is often used to learn an embedding function that captures the semantic differences within a dataset. A key factor in many problem domains is how this embedding generalizes to new classes of data. In observing many triplet selection strategies for Metric Learning, we find that the best performance consistently arises from approaches that focus on a few, well selected triplets.We introduce visualization tools to illustrate how an embedding generalizes beyond measuring accuracy on validation data, and we illustrate the behavior of a range of triplet selection strategies.