Fast Metric Learning For Deep Neural Networks
This work addresses the need for efficient similarity learning in retrieval and classification systems, offering incremental improvements in speed and accuracy.
The paper tackles the problem of learning similarity metrics for information retrieval and machine learning systems by proposing a method that learns target vectors from binary similarity constraints before mapping features to this space, resulting in faster convergence and higher accuracy on most tasks, with concrete improvements demonstrated on multiclass and multi-label datasets.
Similarity metrics are a core component of many information retrieval and machine learning systems. In this work we propose a method capable of learning a similarity metric from data equipped with a binary relation. By considering only the similarity constraints, and initially ignoring the features, we are able to learn target vectors for each instance using one of several appropriately designed loss functions. A regression model can then be constructed that maps novel feature vectors to the same target vector space, resulting in a feature extractor that computes vectors for which a predefined metric is a meaningful measure of similarity. We present results on both multiclass and multi-label classification datasets that demonstrate considerably faster convergence, as well as higher accuracy on the majority of the intrinsic evaluation tasks and all extrinsic evaluation tasks.