CoChat
Sign Up
Cited by 630Advanced Neural Network ApplicationsAdversarial Robustness in Machine LearningDomain Adaptation and Few-Shot Learning

Probability Models for Open Set Recognition

Scheirer, Walter J., Jain, Lalit P., Boult, Terrance E.·2014·IEEE Transactions on Pattern Analysis and Machine Intelligence
DOI 10.1109/tpami.2014.2321392
Open this paper in CoChat — free account

Abstract

Real-world tasks in computer vision often touch upon open set recognition: multi-class recognition with incomplete knowledge of the world and many unknown inputs. Recent work on this problem has proposed a model incorporating an open space risk term to account for the space beyond the reasonable support of known classes. This paper extends the general idea of open space risk limiting classification to accommodate non-linear classifiers in a multiclass setting. We introduce a new open set recognition model called compact abating probability (CAP), where the probability of class membership decreases in value (abates) as points move from known data toward open space. We show that CAP models improve open set recognition for multiple algorithms. Leveraging the CAP formulation, we go on to describe the novel Weibull-calibrated SVM (W-SVM) algorithm, which combines the useful properties of statistical extreme value theory for score calibration with one-class and binary support vector machines. Our experiments show that the W-SVM is significantly better for open set object detection and OCR problems when compared to the state-of-the-art for the same tasks.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/tpami.2014.2321392

Cites

3 of 3 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The current dominant machine learning paradigm makes the closed-world assumption, meaning that classes seen in testing or deployment must have been seen during training.

    “The current dominant machine learning paradigm (ML) makes the closedworld assumption , which means that the classes of objects seen by the system in testing or deployment must have been seen during training”

    From Open-world continual learning: Unifying novelty detection and continual learning (2025) · cites this paper as [59]

    The Introduction (p. 2) says traditional classification considers only known classes, while defining unknown classes as those unseen in training. It calls traditional classification dominant for multi-class computer vision problems, not for machine learning as a whole; the claim broadens that scope. (LLM, conf=0.83)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 20

  1. [1][1] P. L. Bartlett and M. H. Wegkamp. Classification with a reject option using a hinge loss. JMLR , 9:1823-1840, 2008. 3 -
  2. [3][3] C. Bravo, J. L. Lobato, R. Weber, and G. L'Huillier. A hybrid system for probability estimation in multiclass problems combining SVMs and neural networks. In Int. Conf. Hybrid Intel. Sys. , 2008. 3 -
  3. [5][5] C. Chow. On optimum recognition error and reject tradeoff. IEEE Trans. 6. Info. Theory , 16(1):41-46, 1970. 3 -
  4. [7][7] J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei. Imagenet: A large-scale hierarchical image database. In IEEE CVPR , 2009. 2 -
  5. [9][9] R.-E. Fan, K.-W. Chang, C.-J. Hsieh, X.-R. Wang, and C.-J. Lin. LIBLINEAR: A library for large linear classification. JMLR , 9:18711874, 2008. 6 -
  6. [11][11] V. Fragoso and M. Turk. SWIGS: A swift guided sampling method. In IEEE CVPR , June 2013. 3 -
  7. [14][14] J. F. Henriques, J. Carreira, R. Caseiro, and J. Batista. Beyond hard negative mining: Efficient detector learning via block-circulant decomposition. In IEEE ICCV , December 2013. 2 -
  8. [15][15] T. Hofmann, B. Sch¨ olkopf, and A. J. Smola. Kernel methods in machine learning. Annals of Stat. , pp 1171-1220, 2008. 4 -
  9. [16][16] C.-W. Hsu and C.-J. Lin. A comparison of methods for multiclass support vector machines. IEEE T-NN , 13(2):415-425, 2002. 6 -
  10. [18][18] T.-K. Huang, R. C. Weng, and C.-J. Lin. Generalized Bradley-Terry models and multi-class probability estimates. JMLR , 7:85-115, Dec. 2006. 3 -
  11. [20][20] J. A. Lasserre, C. M. Bishop, and T. P. Minka. Principled hybrids of generative and discriminative models. In IEEE CVPR , 2006. 2, 3 -
  12. [21][21] Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner. Gradient-based learning applied to document recognition. Proc. of the IEEE , 86(11):2278-2324, 1998. 2, 7 -
  13. [24][24] W. Scheirer, N. Kumar, P. N. Belhumeur, and T. E. Boult. Multi-attribute spaces: Calibration for attribute fusion and similarity search. In IEEE CVPR , June 2012. 2, 3, 5, 6 -
  14. [25][25] W. Scheirer, A. Rocha, R. Michaels, and T. E. Boult. Meta-recognition: The theory and practice of recognition score analysis. IEEE T-PAMI , 33(8):1689-1695, Aug. 2011. 3, 5 -
  15. [26][26] W. Scheirer, A. Rocha, R. Micheals, and T. Boult. Robust fusion: extreme value theory for recognition score normalization. In ECCV . Springer, 2010. 3 -
  16. [27][27] W. Scheirer, A. Rocha, A. Sapkota, and T. E. Boult. Towards open set recognition. IEEE T-PAMI , 36(7):1757-1772, July 2013. 1, 2, 4, 6, 7 -
  17. [28][28] B. Sch¨ olkopf, J. C. Platt, J. C. Shawe-Taylor, A. J. Smola, and R. C. Williamson. Estimating the support of a high-dimensional distribution. Neural Computation , 13(7):1443-1471, July 2001. 4 -
  18. [29][29] V. Vapnik. The Nature of Statistical Learning Theory, 2nd Edition . Springer, 1998. 2 -
  19. [30][30] Y. Weiss, R. Fergus, and A. Torralba. Multidimensional spectral hashing. In ECCV 2012 , 2012. 4 -
  20. [31][31] R. Zhang and D. Metaxas. RO-SVM: Support vector machine with reject option for image categorization. In BMVC , 2006. 3

Cited by

Sorted by citations

Authors
Scheirer, W. J., Jain, L. P., & Boult, T. E.
Affiliations
  • Harvard University (US)
  • University of Colorado Colorado Springs (US)
ORCID
Venue
IEEE Transactions on Pattern Analysis and Machine Intelligence
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Published
2014-11-01
Volume
36
Issue
11
Pages
2317-2324
ISSN
  • 0162-8828 (ISSN-L)
  • 2160-9292
DOI
10.1109/tpami.2014.2321392
OpenAlex
W2018459374
Cited by
630
References
3
Citations by year
2026: 53 · 2025: 75 · 2024: 75 · 2023: 88 · 2022: 78 · 2021: 70 · 2020: 52 · 2019: 49 · 2018: 35 · 2017: 24 · 2016: 20 · 2015: 8 · 2014: 3
FWCI
15.28
Type
journal article
Language
en
Open access
closed
Licence
ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html (vor)
Locations
  • IEEE Transactions on Pattern Analysis and Machine Intelligence (publishedVersion)
  • PubMed (publishedVersion)
Funders
  • ONR MURI (N00014-08-1-0638)
  • National Science Foundation (IIS-1320956)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
  • Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
  • Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
Keywords
open-set recognition, score calibration, extreme value theory, one-class support vector machine, multi-class classification, unknown class detection, open-set object detection, optical character recognition

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review