 |
ML Newsnew | past | best | rss | submit | about |
| A Deep Metric Learning Reality Check (arxiv.org) | |
| 1 point by wanderer42 on May 11, 2020 | hide | past | pdf | discuss on HN | |
|
In plain words: The study re-checked the experiments behind four years of papers that teach computers to judge how similar two images are. Many used flawed tests, and once corrected, gains over decade-old methods were marginal rather than huge. Abstract · A Metric Learning Reality CheckDeep metric learning papers from the past four years have consistently claimed great advances in accuracy, often more than doubling the performance of decade-old methods. In this paper, we take a closer look at the field to see if this is actually true. We find flaws in the experimental methodology of numerous metric learning papers, and show that the actual improvements over time have been marginal at best. | |
| add comment on HN |