Evaluation of manually created ground truth for multi-view people localization

Ákos Kiss, Tamás Szirányi

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

Abstract

In the field of multi-view people localization, only a few works consider a non-planar ground surface. In this article we introduce a framework for collecting ground truth data in such case, we show characterization of specific errors and introduce a method to automatically merge multiple ground truth data generated by different users to form a more reliable reference ground truth. We use this reference ground truth to evaluate the error rate, the accuracy and the recall of subjects (6 laymen and 3 with domain knowledge). We show that even laymen can work accurately, but even subjects with domain knowledge miss a number of people in a crowded scene. Our findings show that creating ground truth data requires special attention in this field.

Original languageEnglish
Title of host publicationACM International Conference Proceeding Series
DOIs
Publication statusPublished - 2013
EventInternational Workshop on Video and Image Ground Truth in Computer Vision Applications, VIGTA 2013 - St. Petersburg, Russian Federation
Duration: Jul 15 2013Jul 15 2013

Other

OtherInternational Workshop on Video and Image Ground Truth in Computer Vision Applications, VIGTA 2013
CountryRussian Federation
CitySt. Petersburg
Period7/15/137/15/13

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Computer Networks and Communications
  • Computer Vision and Pattern Recognition
  • Software

Cite this

Evaluation of manually created ground truth for multi-view people localization. / Kiss, Ákos; Szirányi, Tamás.

ACM International Conference Proceeding Series. 2013.

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

Kiss, Á & Szirányi, T 2013, Evaluation of manually created ground truth for multi-view people localization. in ACM International Conference Proceeding Series. International Workshop on Video and Image Ground Truth in Computer Vision Applications, VIGTA 2013, St. Petersburg, Russian Federation, 7/15/13. https://doi.org/10.1145/2501105.2501106
Kiss, Ákos ; Szirányi, Tamás. / Evaluation of manually created ground truth for multi-view people localization. ACM International Conference Proceeding Series. 2013.
@inproceedings{b2cd4a3f6f4c41de91152e62d9d125b0,
title = "Evaluation of manually created ground truth for multi-view people localization",
abstract = "In the field of multi-view people localization, only a few works consider a non-planar ground surface. In this article we introduce a framework for collecting ground truth data in such case, we show characterization of specific errors and introduce a method to automatically merge multiple ground truth data generated by different users to form a more reliable reference ground truth. We use this reference ground truth to evaluate the error rate, the accuracy and the recall of subjects (6 laymen and 3 with domain knowledge). We show that even laymen can work accurately, but even subjects with domain knowledge miss a number of people in a crowded scene. Our findings show that creating ground truth data requires special attention in this field.",
author = "{\'A}kos Kiss and Tam{\'a}s Szir{\'a}nyi",
year = "2013",
doi = "10.1145/2501105.2501106",
language = "English",
isbn = "9781450321693",
booktitle = "ACM International Conference Proceeding Series",

}

TY - GEN

T1 - Evaluation of manually created ground truth for multi-view people localization

AU - Kiss, Ákos

AU - Szirányi, Tamás

PY - 2013

Y1 - 2013

N2 - In the field of multi-view people localization, only a few works consider a non-planar ground surface. In this article we introduce a framework for collecting ground truth data in such case, we show characterization of specific errors and introduce a method to automatically merge multiple ground truth data generated by different users to form a more reliable reference ground truth. We use this reference ground truth to evaluate the error rate, the accuracy and the recall of subjects (6 laymen and 3 with domain knowledge). We show that even laymen can work accurately, but even subjects with domain knowledge miss a number of people in a crowded scene. Our findings show that creating ground truth data requires special attention in this field.

AB - In the field of multi-view people localization, only a few works consider a non-planar ground surface. In this article we introduce a framework for collecting ground truth data in such case, we show characterization of specific errors and introduce a method to automatically merge multiple ground truth data generated by different users to form a more reliable reference ground truth. We use this reference ground truth to evaluate the error rate, the accuracy and the recall of subjects (6 laymen and 3 with domain knowledge). We show that even laymen can work accurately, but even subjects with domain knowledge miss a number of people in a crowded scene. Our findings show that creating ground truth data requires special attention in this field.

UR - http://www.scopus.com/inward/record.url?scp=84885202754&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84885202754&partnerID=8YFLogxK

U2 - 10.1145/2501105.2501106

DO - 10.1145/2501105.2501106

M3 - Conference contribution

AN - SCOPUS:84885202754

SN - 9781450321693

BT - ACM International Conference Proceeding Series

ER -