Journal of KIISE (정보과학회 논문지)

Korean Institute of Information Scientists and Engineers (한국정보과학회)
권호 표지
DOI QR코드

DOI QR Code

A Pedestrian Detection Method using Deep Neural Network

심층 신경망을 이용한 보행자 검출 방법

  • 송수호 (선문대학교 컴퓨터융합전자공학과) ;
  • 현훈범 (선문대학교 컴퓨터융합전자공학과) ;
  • 이현 (선문대학교 컴퓨터공학과)
  • Received : 2016.07.06
  • Accepted : 2016.10.24
  • Published : 2017.01.15
⟨ Previous Next ⟩

Abstract

Pedestrian detection, an important component of autonomous driving and driving assistant system, has been extensively studied for many years. In particular, image based pedestrian detection methods such as Hierarchical classifier or HOG and, deep models such as ConvNet are well studied. The evaluation score has increased by the various methods. However, pedestrian detection requires high sensitivity to errors, since small error can lead to life or death problems. Consequently, further reduction in pedestrian detection error rate of autonomous systems is required. We proposed a new method to detect pedestrians and reduce the error rate by using the Faster R-CNN with new developed pedestrian training data sets. Finally, we compared the proposed method with the previous models, in order to show the improvement of our method.

보행자 검출은 수년간 광범위하게 연구된 문제이며, 자율주행 자동차와 운전자 보조시스템에서 매우 중요한 역할을 차지하고 있다. 특히, 계층적 분류기[1]와 Histogram of Gradient[2]특징벡터 등 영상기반의 보행자 검출기법과 ConvNet같이 deep model을 이용하여 검출하는 기법들이 연구되었고 검출성능은 꾸준히 상승하였다. 하지만 보행자 검출은 작은 오차에도 생명과 연관된 문제를 야기할 수 있기 때문에, 자율주행 시스템의 보행자검출 오차율은 더욱 낮출 필요가 있다. 따라서 본 연구에서는 Faster R-CNN 응용 기법에 새로 개발한 데이터 학습 모델을 적용하여 보행자 검출 오류를 줄이는 기법을 제안한다. 그리고 기존에 제안된 모델들과 비교를 통해, 보행자 검출에 있어 제안된 방법의 우수성을 보이고자 한다.

Keywords

References

  1. P. Viola and M. J. Jones, "Rapid Object Detection using a Boosted Cascade of Simple Features," CVPR, pp. 511-518, 2001.
  2. Navneet Dalal and Bill Triggs, "Histograms of oriented gradients for human detection," CVPR, pp. 886-893, 2005.
  3. Q. Zhu, M.-C. Yeh, K.-T. Cheng, and S. Avidan, "Fast human detection using a cascade of histograms of oriented gradients," CVPR, pp. 1491-1498, 2006.
  4. A. Krizhevsky, I. Sutskever, and G. E. Hinton, "ImageNet classification with deep convolutional neural networks," NIPS, pp. 1106-1114, 2012.
  5. P. Dollar, Z. Tu, P. Perona, and S. Belongie, "Integral channel features," BMVC, pp. 91.1-91.11, 2009.
  6. Pedro F Felzenszwalb, Ross B Girshick, David McAllester, and Deva Ramanan, "Object detection with discriminatively trained part-based models," TPAMI, Vol. 32, No. 9, pp. 1627-1645, 2010.
  7. X. Wang, T. X. Han, and S. Yan, "An HOG-LBP human detector with partial occlusion handling," ICCV, pp. 32-39, 2009.
  8. Ross Girshick, Jeff Donahue, Trevor Darrell, and Jagannath Malik, "Rich feature hierarchies for accurate object detection and semantic segmentation," CVPR, pp. 580-587, 2014.
  9. Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun, "Spatial pyramid pooling in deep convolutional networks for visual recognition," ECCV, pp. 346-361. 2014.
  10. Ross Girshick, "Fast r-cnn," ICCV, pp. 1440-1448, 2015.
  11. Pierre Sermanet, David Eigen, Xiang Zhang, Michael Mathieu, Rob Fer-gus, and Yann LeCun, "Overfeat: Integrated recognition, localization and detection using convolutional networks," arXiv:1312.6229, 2013.
  12. S. Ren, K. He, R. Girshick, and J. Sun, "Faster r-cnn: Towards real-time object detection with region proposal networks," NIPS, pp. 91-99, 2015.
  13. C. Szegedy, A. Toshev, and D. Erhan, "Deep neural networks for object detection," NIPS, pp. 2553-2561, 2013.
  14. Y. Jia, E. Shelhamer, J. Donahue, S. Karayev, J. Long, R. Girshick, S. Guadarrama, and T. Darrell, "Caffe: Convolutional architecture for fast feature embedding," ACMMM, pp. 675-678, 2014.
  15. Dollar, P. Wojek, C. Schiele, B. Perona P. "Pedestrian detection: A benchmark," CVPR, pp. 304-311, 2009.
  16. P. Dollar, C. Wojec, B. Schiele, and P. Perona, "Pedestrian detection: An evaluation of the state of the art," PAMI, Vol. 34, No. 4, pp.743-761, Apr. 2012. https://doi.org/10.1109/TPAMI.2011.155
  17. L. Breiman, "Random forests," Machine learning, Vol. 45, No. 1, Oct. 2001.