Efficient optimization for rank-based loss functions

P. Mohapatra, M. Rolinek, C.V. Jawahar, V. Kolmogorov, M.P. Kumar, in:, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, IEEE, 2018, pp. 3693–3701.


Conference Paper | Published | English
Author
; ; ; ;
Department
Abstract
The accuracy of information retrieval systems is often measured using complex loss functions such as the average precision (AP) or the normalized discounted cumulative gain (NDCG). Given a set of positive and negative samples, the parameters of a retrieval system can be estimated by minimizing these loss functions. However, the non-differentiability and non-decomposability of these loss functions does not allow for simple gradient based optimization algorithms. This issue is generally circumvented by either optimizing a structured hinge-loss upper bound to the loss function or by using asymptotic methods like the direct-loss minimization framework. Yet, the high computational complexity of loss-augmented inference, which is necessary for both the frameworks, prohibits its use in large training data sets. To alleviate this deficiency, we present a novel quicksort flavored algorithm for a large class of non-decomposable loss functions. We provide a complete characterization of the loss functions that are amenable to our algorithm, and show that it includes both AP and NDCG based loss functions. Furthermore, we prove that no comparison based algorithm can improve upon the computational complexity of our approach asymptotically. We demonstrate the effectiveness of our approach in the context of optimizing the structured hinge loss upper bound of AP and NDCG loss for learning models for a variety of vision tasks. We show that our approach provides significantly better results than simpler decomposable loss functions, while requiring a comparable training time.
Publishing Year
Date Published
2018-06-28
Proceedings Title
2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
Page
3693-3701
Conference
CVPR: Conference on Computer Vision and Pattern Recognition
Conference Location
Salt Lake City, UT, USA
Conference Date
2018-06-18 – 2018-06-22
IST-REx-ID

Cite this

Mohapatra P, Rolinek M, Jawahar CV, Kolmogorov V, Kumar MP. Efficient optimization for rank-based loss functions. In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE; 2018:3693-3701. doi:10.1109/cvpr.2018.00389
Mohapatra, P., Rolinek, M., Jawahar, C. V., Kolmogorov, V., & Kumar, M. P. (2018). Efficient optimization for rank-based loss functions. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 3693–3701). Salt Lake City, UT, USA: IEEE. https://doi.org/10.1109/cvpr.2018.00389
Mohapatra, Pritish, Michal Rolinek, C V Jawahar, Vladimir Kolmogorov, and M Pawan Kumar. “Efficient Optimization for Rank-Based Loss Functions.” In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 3693–3701. IEEE, 2018. https://doi.org/10.1109/cvpr.2018.00389.
P. Mohapatra, M. Rolinek, C. V. Jawahar, V. Kolmogorov, and M. P. Kumar, “Efficient optimization for rank-based loss functions,” in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 2018, pp. 3693–3701.
Mohapatra P, Rolinek M, Jawahar CV, Kolmogorov V, Kumar MP. 2018. Efficient optimization for rank-based loss functions. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. CVPR: Conference on Computer Vision and Pattern Recognition 3693–3701.
Mohapatra, Pritish, et al. “Efficient Optimization for Rank-Based Loss Functions.” 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, IEEE, 2018, pp. 3693–701, doi:10.1109/cvpr.2018.00389.

Link(s) to Main File(s)
Access Level
OA Open Access

Export

Marked Publications

Open Data IST Research Explorer

Sources

arXiv 1604.08269

Search this title in

Google Scholar
ISBN Search