Advertisement

Abstract

This paper presents an Ensemble R-FCN framework for object detection. Specifically, we mainly make three contributions to our detection framework: (1) we augment the training images for R-FCN when facing the limited training samples and small object. (2) We further introduce several enhancement schemes to improve the performance of the single R-FCN. (3) An ensemble R-FCN is proposed to make our detection system more robust by combining different feature extractors and multi-scale inference. Experimental results demonstrate the advantages of the proposed method. Especially, our method achieved the performance of AP score 0.829 which ranked No. 1 among over 360 teams in Ucar Self-driving deep learning Competition.

Keywords

Object detection R-FCN Self-driving Deep learning 

References

  1. 1.
    Lowe, D.: Distinctive image features from scale-invariant key-points. In: IJCV (2004) Google Scholar
  2. 2.
    Dalal, N., Triggs, N.: Histograms of oriented gradients for human detection. In: CVPR (2005)Google Scholar
  3. 3.
    Hoiem, D., Chodpathumwan, Y., Dai, Q.: Diagnosing error in object detectors. In: ECCV 2012 (2012)Google Scholar
  4. 4.
    Girshick, R., Donahue, J., Darrell, T., Malik, J.: Rich feature hierarchies for accurate object detection and semantic segmentation. In: CVPR (2014)Google Scholar
  5. 5.
    Van de Sande, K.E., Uijlings, J.R., Gevers, T., Smeulders, A.W.: Segmentation as selective search for object recognition. In: ICCV (2011)Google Scholar
  6. 6.
    He, K., Zhang, X., Ren, S., Sun, J.: Spatial pyramid pooling in deep convolutional networks for visual recognition. In: ECCV (2014)Google Scholar
  7. 7.
    Girshick, R.: Fast R-CNN. In: ICCV (2015)Google Scholar
  8. 8.
    Ren, S., He, K., Girshick, R., Sun, J.: Faster R-CNN: towards real-time object detection with region proposal networks. In: NIPS (2015)Google Scholar
  9. 9.
    Dai, J., Li, Y., He, K., Sun, J.: R-FCN: object detection via region-based fully convolutional networks. arXiv preprint arXiv:1605.06409 (2016)
  10. 10.
    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: CVPR (2016)Google Scholar
  11. 11.
    Li, D., Li, J., Nie, B., Sun, S.: Deconvolution single shot multibox detector for supermarket commodity detection and classification. In: ICDIP (2017) Google Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2018

Authors and Affiliations

  1. 1.Computer Science and EngineeringNanjing University of Science and TechnologyNanjingChina
  2. 2.School of Computer and SoftwareNanjing University of Information Science and TechnologyNanjingChina

Personalised recommendations