Catching Image Retrieval Generalization
This work addresses a fundamental evaluation problem in image retrieval for researchers and practitioners, offering a more reliable metric for assessing model generalization.
The authors identified that the Recall@K metric's dependence on dataset class count limits its ability to estimate generalization in image retrieval, and they proposed a new metric that addresses this issue, providing new insights into deep metric learning generalization.
The concepts of overfitting and generalization are vital for evaluating machine learning models. In this work, we show that the popular Recall@K metric depends on the number of classes in the dataset, which limits its ability to estimate generalization. To fix this issue, we propose a new metric, which measures retrieval performance, and, unlike Recall@K, estimates generalization. We apply the proposed metric to popular image retrieval methods and provide new insights about deep metric learning generalization.