Skip to main content
SpringerLink
Log in

EnMS: early non-maxima suppression

Speeding up pattern localization and other tasks

  • Theoretical Advances
  • Published:
Pattern Analysis and Applications Aims and scope Submit manuscript

Abstract

Detection of objects in images using statistical classifiers is a well studied and documented technique. Different applications of such detectors often require selection of the image position with the highest response of the detector—they perform non-maxima suppression. This article introduces the concept of early non-maxima suppression, which aims to reduce necessary computations by making the non-maxima suppression decision early based on incomplete information provided by a partially evaluated classifier. We show that the error of one such speculative decision with respect to a decision made based on response of the complete classifier can be estimated by collecting statistics on unlabeled data. The article then considers a sequential strategy of multiple early non-maxima suppression tests which follows the structure of soft-cascade detectors commonly used for object detection. We also show that an optimal (fastest for requested error rate) suppression strategy can be created by a novel variant of Wald’s sequential probability ratio test (SPRT) which we call the conditioned SPRT (CSPRT). Experimental results show that the early non-maxima suppression significantly reduces amount of computation in the case of object localization while the error rates are limited to low predefined values. The proposed approach notably outperforms the state-of-the-art detectors based on WaldBoost. The potential applications of the early non-maxima suppression approach are not limited to object localization and could be applied wherever the goal is to find the strongest response of a classifier among a set of classified samples.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7

Similar content being viewed by others

References

  1. Bishop CM (2006) Pattern recognition and machine learning (information science and statistics). Springer, New York

  2. Boom BJ, van Rootseler RTA, Veldhuis RNJ (2007) Investigating the boosting framework for face recognition. In: Veldhuis RNJ, Cronie HS (eds) Proceedings of the 28th symposium on information theory in the Benelux, Eindhoven, May 2007. Werkgemeenschap voor Informatie- en Communicatietechniek, pp 189–196

  3. Bourdev L, Brandt J (2005) Robust object detection via soft cascade. In: CVPR

  4. Brubaker CS, Mullin MD, Rehg JM (2006) Towards optimal training of cascaded detectors. In: ECCV06, pp 325–337

  5. Cao Y, Xu J, Liu T-Y, Li H, Huang Y, Hon H-W (2006) Adapting ranking SVM to document retrieval. In: SIGIR ’06: proceedings of the 29th annual international ACM SIGIR conference on research and development in information retrieval. ACM, New York (pp 186–193)

  6. Cha Z, Viola P (2007) Multiple-instance pruning for learning efficient cascade detectors. In: NIPS

  7. Grabner H, Grabner M, Bischof H (2006) Real-time tracking via on-line boosting, vol 1, p 47

  8. Grabner H, Šochman J, Bischof H, Matas JG (2008) Training sequential on-line boosting classifier for visual tracking. In: ICPR08, international conference on pattern recognition, pp 1–4

  9. Herout A, Hradiš M, Jošth R, Juranek R, Zemčik P (2008) “Local rank differences” image faeture implemented on GPU. In: Advanced concepts for intelligent vision systems. Lecture notes in computer science, pp 170–181

  10. Hradiš M, Herout A, Zemčik P (2008) Local rank patterns—novel features for rapid object detection. In: Proceedings of international conference on computer vision and graphics 2008. Lecture notes in computer science, pp 1–12

  11. Huang C, Ai HZ, Li Y, Lao SH (2007) High-performance rotation invariant multiview face detection. PAMI 29(4):671–686

    Article  Google Scholar 

  12. Jesorsky O, Kirchberg KJ, Frischholz R (2001) Robust face detection using the Hausdorff distance. In: AVBPA ’01: proceedings of the third international conference on audio- and video-based biometric person authentication. Springer, London, pp 90–95

  13. Messer K, Matas J, Kittler J, Luettin J, Maiter G (1999) XM2VTSDB: the extended M2VTS database. In: Second international conference on audio and video-based biometric person authentication, pp 72–77

  14. Mutch J, Lowe DG (2006) Multiclass object recognition with sparse, localized features. In: 2006 IEEE Computer Society conference on computer vision and pattern recognition, vol 1, pp 11–18

  15. Schapire RE, Singer Y (1999) Improved boosting algorithms using confidence-rated predictions. Mach Learn 37:297–336

    Google Scholar 

  16. Schneiderman H (2004) Feature-centric evaluation for efficient cascaded object detection. In: IEEE Computer Society conference on computer vision and pattern recognition, vol 2, pp 29–36

  17. Šochman J, Matas J (2005) WaldBoost—learning for time constrained sequential detection. In: CVPR ’05: proceedings of the 2005 IEEE Computer Society conference on computer vision and pattern recognition (CVPR’05), vol 2. IEEE Computer Society, Washington, DC. pp 150–156

  18. Šochman J, Matas J (2009) Learning fast emulators of binary decision processes. Int J Comput Vision 83(2):149–163

    Google Scholar 

  19. Viola P, Jones M (2001) Rapid object detection using a boosted cascade of simple features. In: IEEE Computer Society conference on computer vision and pattern recognition, vol 1, p 511

  20. Wald A (1945) Sequential tests of statistical hypotheses. Ann Math Stat 16(2):117–186

    Article  MathSciNet  MATH  Google Scholar 

Download references

Acknowledgments

This work was supported by the Ministry of Education, Youth and Sports of the Czech Republic under the research program LC-06008 (Center for Computer Graphics) and by the research project “Security-Oriented Research in Information Technology” CEZMSMT, MSM0021630528. The research presented in this article acknowledges the use of the Extended Multimodal Face Database and associated documentation. Further details of this software can be found in [13].

Author information

Authors and Affiliations

  1. Graph@FIT, Brno University of Technology, Bozetechova 2, 616 00, Brno, Czech Republic

    Adam Herout, Michal Hradiš & Pavel Zemčík

Authors
  1. Adam Herout
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Michal Hradiš
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Pavel Zemčík
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Adam Herout.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Herout, A., Hradiš, M. & Zemčík, P. EnMS: early non-maxima suppression. Pattern Anal Applic 15, 121–132 (2012). https://doi.org/10.1007/s10044-011-0213-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10044-011-0213-2

Keywords

Search

Navigation