Black-Box Optimization of Object Detector Scales
- Object detectors have improved considerably in the last years by using advanced CNN architectures. However, many detector hyper-parameters are generally manually tuned, or they are used with values set by the detector authors. Automatic Hyper-parameter optimization has not been explored in improving CNN-based object detectors hyper-parameters. In this work, we propose the use of Black-box optimization methods to tune the prior/default box scales in Faster R-CNN and SSD, using Bayesian Optimization, SMAC, and CMA-ES. We show that by tuning the input image size and prior box anchor scale on Faster R-CNN mAP increases by 2% on PASCAL VOC 2007, and by 3% with SSD. On the COCO dataset with SSD there are mAP improvement in the medium and large objects, but mAP decreases by 1% in small objects. We also perform a regression analysis to find the significant hyper-parameters to tune.
| Document Type: | Preprint |
|---|---|
| Language: | English |
| Author: | Mohandass Muthuraja, Octavio Arriaga, Paul Plöger, Frank Kirchner, Matias Valdenegro-Toro |
| Number of pages: | 17 |
| DOI: | https://doi.org/10.48550/arXiv.2010.15823 |
| ArXiv Id: | http://arxiv.org/abs/2010.15823 |
| Publisher: | arXiv |
| Date of first publication: | 2020/10/29 |
| Keyword: | Black-Box Optimization; Hyper-parameter Tuning; Scale Tuning; object detection |
| Departments, institutes and facilities: | Fachbereich Informatik |
| Dewey Decimal Classification (DDC): | 0 Informatik, Informationswissenschaft, allgemeine Werke / 00 Informatik, Wissen, Systeme / 004 Datenverarbeitung; Informatik |
| Entry in this database: | 2020/11/04 |
