Active constrained clustering via non-iterative uncertainty sampling

Panagiotis Stanitsas, Anoop Cherian, Vassilios Morellas, Nikolaos P Papanikolopoulos

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

Active Constraint Learning (ACL) is continuously gaining popularity in the area of constrained clustering due to its ability to achieve performance gains via incorporating minimal feedback from a human annotator for selected instances. For constrained clustering algorithms, such instances are integrated in the form of Must-Link (ML) and Cannot-Link (CL) constraints. Existing iterative uncertainty reduction schemes, introduce high computational burden particularly when they process larger datasets that are usually present in computer vision and visual learning applications. For scenarios that multiple agents (i.e., robots) require user feedback for performing recognition tasks, minimizing the interaction between the user and the agents, without compromising performance, is an essential task. In this study, a non-iterative ACL scheme with proven performance benefits is presented. We select to demonstrate the effectiveness of our methodology by building on the well known K-Means algorithm for clustering; one can easily extend it to alternative clustering schemes. The proposed methodology introduces the use of the Silhouette values, conventionally used for measuring clustering performance, in order to rank the degree of information content of the various samples. In addition, an efficient greedy selection scheme was devised for selecting the most informative samples for human annotation. To the best of our knowledge, this is the first active constrained clustering methodology with the ability to process computer vision datasets that this study targets. Performance results are shown on various computer vision benchmarks and support the merits of adopting the proposed scheme.

Original languageEnglish (US)
Title of host publicationIROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages4027-4033
Number of pages7
ISBN (Electronic)9781509037629
DOIs
StatePublished - Nov 28 2016
Event2016 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2016 - Daejeon, Korea, Republic of
Duration: Oct 9 2016Oct 14 2016

Publication series

NameIEEE International Conference on Intelligent Robots and Systems
Volume2016-November
ISSN (Print)2153-0858
ISSN (Electronic)2153-0866

Other

Other2016 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2016
CountryKorea, Republic of
CityDaejeon
Period10/9/1610/14/16

Fingerprint

Computer vision
Sampling
Feedback
Clustering algorithms
Robots
Uncertainty

Keywords

  • Active constrained clustering
  • Image clustering uncertainty management
  • Visual learning

Cite this

Stanitsas, P., Cherian, A., Morellas, V., & Papanikolopoulos, N. P. (2016). Active constrained clustering via non-iterative uncertainty sampling. In IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (pp. 4027-4033). [7759593] (IEEE International Conference on Intelligent Robots and Systems; Vol. 2016-November). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IROS.2016.7759593

Active constrained clustering via non-iterative uncertainty sampling. / Stanitsas, Panagiotis; Cherian, Anoop; Morellas, Vassilios; Papanikolopoulos, Nikolaos P.

IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems. Institute of Electrical and Electronics Engineers Inc., 2016. p. 4027-4033 7759593 (IEEE International Conference on Intelligent Robots and Systems; Vol. 2016-November).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Stanitsas, P, Cherian, A, Morellas, V & Papanikolopoulos, NP 2016, Active constrained clustering via non-iterative uncertainty sampling. in IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems., 7759593, IEEE International Conference on Intelligent Robots and Systems, vol. 2016-November, Institute of Electrical and Electronics Engineers Inc., pp. 4027-4033, 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2016, Daejeon, Korea, Republic of, 10/9/16. https://doi.org/10.1109/IROS.2016.7759593
Stanitsas P, Cherian A, Morellas V, Papanikolopoulos NP. Active constrained clustering via non-iterative uncertainty sampling. In IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems. Institute of Electrical and Electronics Engineers Inc. 2016. p. 4027-4033. 7759593. (IEEE International Conference on Intelligent Robots and Systems). https://doi.org/10.1109/IROS.2016.7759593
Stanitsas, Panagiotis ; Cherian, Anoop ; Morellas, Vassilios ; Papanikolopoulos, Nikolaos P. / Active constrained clustering via non-iterative uncertainty sampling. IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems. Institute of Electrical and Electronics Engineers Inc., 2016. pp. 4027-4033 (IEEE International Conference on Intelligent Robots and Systems).
@inproceedings{d2f6e000705f414bae29349c957ddc60,
title = "Active constrained clustering via non-iterative uncertainty sampling",
abstract = "Active Constraint Learning (ACL) is continuously gaining popularity in the area of constrained clustering due to its ability to achieve performance gains via incorporating minimal feedback from a human annotator for selected instances. For constrained clustering algorithms, such instances are integrated in the form of Must-Link (ML) and Cannot-Link (CL) constraints. Existing iterative uncertainty reduction schemes, introduce high computational burden particularly when they process larger datasets that are usually present in computer vision and visual learning applications. For scenarios that multiple agents (i.e., robots) require user feedback for performing recognition tasks, minimizing the interaction between the user and the agents, without compromising performance, is an essential task. In this study, a non-iterative ACL scheme with proven performance benefits is presented. We select to demonstrate the effectiveness of our methodology by building on the well known K-Means algorithm for clustering; one can easily extend it to alternative clustering schemes. The proposed methodology introduces the use of the Silhouette values, conventionally used for measuring clustering performance, in order to rank the degree of information content of the various samples. In addition, an efficient greedy selection scheme was devised for selecting the most informative samples for human annotation. To the best of our knowledge, this is the first active constrained clustering methodology with the ability to process computer vision datasets that this study targets. Performance results are shown on various computer vision benchmarks and support the merits of adopting the proposed scheme.",
keywords = "Active constrained clustering, Image clustering uncertainty management, Visual learning",
author = "Panagiotis Stanitsas and Anoop Cherian and Vassilios Morellas and Papanikolopoulos, {Nikolaos P}",
year = "2016",
month = "11",
day = "28",
doi = "10.1109/IROS.2016.7759593",
language = "English (US)",
series = "IEEE International Conference on Intelligent Robots and Systems",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "4027--4033",
booktitle = "IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems",

}

TY - GEN

T1 - Active constrained clustering via non-iterative uncertainty sampling

AU - Stanitsas, Panagiotis

AU - Cherian, Anoop

AU - Morellas, Vassilios

AU - Papanikolopoulos, Nikolaos P

PY - 2016/11/28

Y1 - 2016/11/28

N2 - Active Constraint Learning (ACL) is continuously gaining popularity in the area of constrained clustering due to its ability to achieve performance gains via incorporating minimal feedback from a human annotator for selected instances. For constrained clustering algorithms, such instances are integrated in the form of Must-Link (ML) and Cannot-Link (CL) constraints. Existing iterative uncertainty reduction schemes, introduce high computational burden particularly when they process larger datasets that are usually present in computer vision and visual learning applications. For scenarios that multiple agents (i.e., robots) require user feedback for performing recognition tasks, minimizing the interaction between the user and the agents, without compromising performance, is an essential task. In this study, a non-iterative ACL scheme with proven performance benefits is presented. We select to demonstrate the effectiveness of our methodology by building on the well known K-Means algorithm for clustering; one can easily extend it to alternative clustering schemes. The proposed methodology introduces the use of the Silhouette values, conventionally used for measuring clustering performance, in order to rank the degree of information content of the various samples. In addition, an efficient greedy selection scheme was devised for selecting the most informative samples for human annotation. To the best of our knowledge, this is the first active constrained clustering methodology with the ability to process computer vision datasets that this study targets. Performance results are shown on various computer vision benchmarks and support the merits of adopting the proposed scheme.

AB - Active Constraint Learning (ACL) is continuously gaining popularity in the area of constrained clustering due to its ability to achieve performance gains via incorporating minimal feedback from a human annotator for selected instances. For constrained clustering algorithms, such instances are integrated in the form of Must-Link (ML) and Cannot-Link (CL) constraints. Existing iterative uncertainty reduction schemes, introduce high computational burden particularly when they process larger datasets that are usually present in computer vision and visual learning applications. For scenarios that multiple agents (i.e., robots) require user feedback for performing recognition tasks, minimizing the interaction between the user and the agents, without compromising performance, is an essential task. In this study, a non-iterative ACL scheme with proven performance benefits is presented. We select to demonstrate the effectiveness of our methodology by building on the well known K-Means algorithm for clustering; one can easily extend it to alternative clustering schemes. The proposed methodology introduces the use of the Silhouette values, conventionally used for measuring clustering performance, in order to rank the degree of information content of the various samples. In addition, an efficient greedy selection scheme was devised for selecting the most informative samples for human annotation. To the best of our knowledge, this is the first active constrained clustering methodology with the ability to process computer vision datasets that this study targets. Performance results are shown on various computer vision benchmarks and support the merits of adopting the proposed scheme.

KW - Active constrained clustering

KW - Image clustering uncertainty management

KW - Visual learning

UR - http://www.scopus.com/inward/record.url?scp=85006507955&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85006507955&partnerID=8YFLogxK

U2 - 10.1109/IROS.2016.7759593

DO - 10.1109/IROS.2016.7759593

M3 - Conference contribution

T3 - IEEE International Conference on Intelligent Robots and Systems

SP - 4027

EP - 4033

BT - IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems

PB - Institute of Electrical and Electronics Engineers Inc.

ER -