Header logo is

Optimizing Rank-based Metrics with Blackbox Differentiation


Conference Paper


Rank-based metrics are some of the most widely used criteria for performance evaluation of computer vision models. Despite years of effort, direct optimization for these metrics remains a challenge due to their non-differentiable and non-decomposable nature. We present an efficient, theoretically sound, and general method for differentiating rank-based metrics with mini-batch gradient descent. In addition, we address optimization instability and sparsity of the supervision signal that both arise from using rank-based metrics as optimization targets. Resulting losses based on recall and Average Precision are applied to image retrieval and object detection tasks. We obtain performance that is competitive with state-of-the-art on standard image retrieval datasets and consistently improve performance of near state-of-the-art object detectors.

Author(s): Michal Rolinek and Vít Musil and Anselm Paulus and Marin Vlastelica and Claudio Michaelis and Georg Martius
Book Title: Proceedings IEEE Conf. on Computer Vision and Pattern Recognition (CVPR)
Pages: 7620-7630
Year: 2020

Department(s): Autonomous Learning
Research Project(s): Differentiation of Blackbox Combinatorial Solvers
Bibtex Type: Conference Paper (inproceedings)
Paper Type: Conference

Event Name: IEEE International Conference on Computer Vision and Pattern Recognition (CVPR) 2020
Event Place: Seattle, USA

Note: Best paper nomination

Links: Paper @ CVPR
Long Oral
Short Oral


  title = {Optimizing Rank-based Metrics with Blackbox Differentiation},
  author = {Rolinek, Michal and Musil, Vít and Paulus, Anselm and Vlastelica, Marin and Michaelis, Claudio and Martius, Georg},
  booktitle = { Proceedings IEEE Conf. on Computer Vision and Pattern Recognition (CVPR)},
  pages = {7620-7630},
  year = {2020},
  note = {Best paper nomination}