Bucketed Ranking-Based Losses for Efficient Training of Object Detectors

Date

2025-01-01

Author

Yavuz, Feyza
Cam, Baris Can
Doğan, Adnan Harun
Oksuz, Kemal
Akbaş, Emre
KALKAN, SİNAN

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

18
views

0
downloads

Cite This

Ranking-based loss functions, such as Average Precision Loss and Rank&Sort Loss, outperform widely used score-based losses in object detection. These loss functions better align with the evaluation criteria, have fewer hyperparameters, and offer robustness against the imbalance between positive and negative classes. However, they require pairwise comparisons among P positive and N negative predictions, introducing a time complexity of O(PN), which is prohibitive since N is often large (e.g., 108 in ATSS). Despite their advantages, the widespread adoption of ranking-based losses has been hindered by their high time and space complexities. In this paper, we focus on improving the efficiency of ranking-based loss functions. To this end, we propose Bucketed Ranking-based Losses which group negative predictions into B buckets (B≪N) in order to reduce the number of pairwise comparisons so that time complexity can be reduced. Our method enhances the time complexity, reducing it to O(max(Nlog(N),P2)). To validate our method and show its generality, we conducted experiments on 2 different tasks, 3 different datasets, 7 different detectors. We show that Bucketed Ranking-based (BR) Losses yield the same accuracy with the unbucketed versions and provide 2× faster training on average. We also train, for the first time, transformer-based object detectors using ranking-based losses, thanks to the efficiency of our BR. When we train CoDETR, a state-of-the-art transformer-based object detector, using our BR Loss, we consistently outperform its original results over several different backbones. Code will be released.

Subject Keywords

Efficient Ranking Losses, Object Detection, Ranking Losses

URI

https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85210839819&origin=inward
https://hdl.handle.net/11511/113154

Collections

Department of Computer Engineering, Conference / Seminar