IS THE GAME WORTH THE CANDLE? - Evaluation of OpenCL for Object Detection Algorithm Optimization

Floris De Smedt, Lars Struyf, Sander Beckers, Joost Vennekens, Gorik De Samblanx, Toon Goedemé

Abstract

In this paper we present out experiences with the implementation of an object detector using OpenCL. With this implementation we fullfil the need for fast and robust object detection, necessary in many applications in multiple domains (surveillance, traffic, image retrieval, ...). The algorithm lends itself to be implemented in a parallel way. We exploit this opportunity by implementing it on a GPU. For this implementation, we have choosen to use the OpenCL programming language, since this allows for scalability to more performant and different types of hardware, with minimal changes to the implementation. We will discuss how the parallelization is done, and discuss the challenges we met. We will also discuss the experimental timing results we achieved and evaluate the ease-of-use of OpenCL.

References

  1. Benedict, G. R., David, K., Perhaad, M., and Dana, S. (2011). Heterogeneous Computing with OpenCL. Morgan Kaupmann.
  2. Dalal, N. and Triggs, B. (2005). Histograms of oriented gradients for human detection. In International Conf. on CVPR, volume 2, pages 886-893.
  3. Felzenszwalb, P., Girschick, R., and McAllester, D. (2010a). Cascade object detection with deformable part models. In Proc. of the IEEE Conf. on CVPR.
  4. Felzenszwalb, P., Girschick, R., McAllester, D., and Ramanan, D. (2010b). Object detection with discriminatively trained part based models. IEEE Transactions on Pattern Analysis and Machine Intelligence, 32(9).
  5. Felzenszwalb, P., McAllester, D., and Ramanan, D. (2008). A discriminatively trained, multiscale, deformable part model. In Proc. of the IEEE Conf. on CVPR.
  6. Felzenszwalb, P. F., Girshick, R. B., and McAllester, D. (2010c). Discriminatively trained deformable part models, release 4. http://people.cs.uchicago.edu/ pff/latent-release4/.
  7. Gall, J., Yao, A., Razavi, N., Van Gool, L., and Lempitsky, V. (2011). Hough forests for object detection, tracking, and action recognition. In IEEE Transactions on Pattern Analysis and Machine Intelligence.
  8. Group, K. (2011). Opencl - the open standard for parallel programming of heterogeneous systems. http://www.khronos.org/opencl/.
  9. Leibe, B., Leonardis, A., and B.Schiele (2004). Combined object categorization and segmentation with an implicit shape model. In ECCV'04 Workshop on Statistical Learning in Computer Vision.
  10. Lowe, D. G. (2004). Distinctive image features from scaleinvariant keypoints. International Journal of Computer Vision.
  11. (2010). Pets 2010 benchmark http://www.cvg.rdg.ac.uk/PETS2010/a.html.
  12. Tsuchiyama, R., Nakamura, T., Lizuka, T., Asahara, A., and Miki, S. (2009). The OpenCL Programming book. Fixstars.
  13. Van Beeck, K., De Smedt, F., Beckers, S., Struyf, L., Vennekens, J., De Samblanx, G., Goedemé, T., and Tuytelaars, T. (2011). Towards robust automatic detection of vulnerable road users: Monocular pedestrian tracking from a moving vehicle,. In Proc. of ATINER 7th Annual International Conf. on Computer Science and Information Systems.
  14. Viola, P. and Jones, M. (2001). Rapid object detection using a boosted cascade of simple features. In Proc. of the IEEE Conf. on CVPR.
Download


Paper Citation


in Harvard Style

De Smedt F., Struyf L., Beckers S., Vennekens J., De Samblanx G. and Goedemé T. (2012). IS THE GAME WORTH THE CANDLE? - Evaluation of OpenCL for Object Detection Algorithm Optimization . In Proceedings of the 2nd International Conference on Pervasive Embedded Computing and Communication Systems - Volume 1: PECCS, ISBN 978-989-8565-00-6, pages 284-291. DOI: 10.5220/0003821002840291


in Bibtex Style

@conference{peccs12,
author={Floris De Smedt and Lars Struyf and Sander Beckers and Joost Vennekens and Gorik De Samblanx and Toon Goedemé},
title={IS THE GAME WORTH THE CANDLE? - Evaluation of OpenCL for Object Detection Algorithm Optimization},
booktitle={Proceedings of the 2nd International Conference on Pervasive Embedded Computing and Communication Systems - Volume 1: PECCS,},
year={2012},
pages={284-291},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0003821002840291},
isbn={978-989-8565-00-6},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 2nd International Conference on Pervasive Embedded Computing and Communication Systems - Volume 1: PECCS,
TI - IS THE GAME WORTH THE CANDLE? - Evaluation of OpenCL for Object Detection Algorithm Optimization
SN - 978-989-8565-00-6
AU - De Smedt F.
AU - Struyf L.
AU - Beckers S.
AU - Vennekens J.
AU - De Samblanx G.
AU - Goedemé T.
PY - 2012
SP - 284
EP - 291
DO - 10.5220/0003821002840291