Traffic Light Detection: A Learning Algorithm and Evaluations on Challenging Dataset

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

29 Citationer (Scopus)
822 Downloads (Pure)

Resumé

Traffic light recognition (TLR) is an integral part of any intelligent vehicle, which must function in the existing infrastructure. Pedestrian and sign detection have recently seen great improvements due to the introduction of learning based detectors using integral channel features. A similar push have not yet been seen for the detection sub-problem of TLR, where detection is dominated by methods based on heuristic models.
Evaluation of existing systems is currently limited primarily to small local datasets. In order to provide a common basis for comparing future TLR research an extensive public database is collected based on footage from US roads. The database consists of both test and training data, totaling 46,418 frames and 112,971 annotated traffic lights, captured in continuous sequences under a varying light and weather conditions.
The learning based detector achieves an AUC of 0.4 and 0.32 for day sequence 1 and 2, respectively, which is more than an order of magnitude better than the two heuristic model-based detectors.
OriginalsprogEngelsk
Titel2015 IEEE 18th International Conference on Intelligent Transportation Systems (ITSC 2015) : Proceedings
ForlagIEEE
Publikationsdato2015
Sider2341-2345
Artikelnummer7313470
ISBN (Trykt)978-1-4673-6595-6
DOI
StatusUdgivet - 2015
Begivenhed18th IEEE Intelligent Transportation Systems Conference - Las Palmas de Gran Canaria, Spanien
Varighed: 15 sep. 201518 sep. 2015
Konferencens nummer: 18
http://www.ieee.org/conferences_events/conferences/conferencedetails/index.html?Conf_ID=30712

Konference

Konference18th IEEE Intelligent Transportation Systems Conference
Nummer18
LandSpanien
ByLas Palmas de Gran Canaria
Periode15/09/201518/09/2015
Internetadresse
NavnI E E E Intelligent Transportation Systems Conference. Proceedings

Fingerprint

Telecommunication traffic
Learning algorithms
Detectors
Intelligent vehicle highway systems

Citer dette

Philipsen, M. P., Jensen, M. B., Møgelmose, A., Moeslund, T. B., & Trivedi, M. M. (2015). Traffic Light Detection: A Learning Algorithm and Evaluations on Challenging Dataset. I 2015 IEEE 18th International Conference on Intelligent Transportation Systems (ITSC 2015): Proceedings (s. 2341-2345). [7313470] IEEE. I E E E Intelligent Transportation Systems Conference. Proceedings https://doi.org/10.1109/ITSC.2015.378
Philipsen, Mark Philip ; Jensen, Morten Bornø ; Møgelmose, Andreas ; Moeslund, Thomas B. ; Trivedi, Mohan M. / Traffic Light Detection : A Learning Algorithm and Evaluations on Challenging Dataset. 2015 IEEE 18th International Conference on Intelligent Transportation Systems (ITSC 2015): Proceedings. IEEE, 2015. s. 2341-2345 (I E E E Intelligent Transportation Systems Conference. Proceedings).
@inproceedings{e86c6162493e4714a3fc2e2cf50b7056,
title = "Traffic Light Detection: A Learning Algorithm and Evaluations on Challenging Dataset",
abstract = "Traffic light recognition (TLR) is an integral part of any intelligent vehicle, which must function in the existing infrastructure. Pedestrian and sign detection have recently seen great improvements due to the introduction of learning based detectors using integral channel features. A similar push have not yet been seen for the detection sub-problem of TLR, where detection is dominated by methods based on heuristic models.Evaluation of existing systems is currently limited primarily to small local datasets. In order to provide a common basis for comparing future TLR research an extensive public database is collected based on footage from US roads. The database consists of both test and training data, totaling 46,418 frames and 112,971 annotated traffic lights, captured in continuous sequences under a varying light and weather conditions.The learning based detector achieves an AUC of 0.4 and 0.32 for day sequence 1 and 2, respectively, which is more than an order of magnitude better than the two heuristic model-based detectors.",
author = "Philipsen, {Mark Philip} and Jensen, {Morten Born{\o}} and Andreas M{\o}gelmose and Moeslund, {Thomas B.} and Trivedi, {Mohan M.}",
year = "2015",
doi = "10.1109/ITSC.2015.378",
language = "English",
isbn = "978-1-4673-6595-6",
series = "I E E E Intelligent Transportation Systems Conference. Proceedings",
pages = "2341--2345",
booktitle = "2015 IEEE 18th International Conference on Intelligent Transportation Systems (ITSC 2015)",
publisher = "IEEE",
address = "United States",

}

Philipsen, MP, Jensen, MB, Møgelmose, A, Moeslund, TB & Trivedi, MM 2015, Traffic Light Detection: A Learning Algorithm and Evaluations on Challenging Dataset. i 2015 IEEE 18th International Conference on Intelligent Transportation Systems (ITSC 2015): Proceedings., 7313470, IEEE, I E E E Intelligent Transportation Systems Conference. Proceedings, s. 2341-2345, 18th IEEE Intelligent Transportation Systems Conference, Las Palmas de Gran Canaria, Spanien, 15/09/2015. https://doi.org/10.1109/ITSC.2015.378

Traffic Light Detection : A Learning Algorithm and Evaluations on Challenging Dataset. / Philipsen, Mark Philip; Jensen, Morten Bornø; Møgelmose, Andreas; Moeslund, Thomas B.; Trivedi, Mohan M.

2015 IEEE 18th International Conference on Intelligent Transportation Systems (ITSC 2015): Proceedings. IEEE, 2015. s. 2341-2345 7313470 (I E E E Intelligent Transportation Systems Conference. Proceedings).

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

TY - GEN

T1 - Traffic Light Detection

T2 - A Learning Algorithm and Evaluations on Challenging Dataset

AU - Philipsen, Mark Philip

AU - Jensen, Morten Bornø

AU - Møgelmose, Andreas

AU - Moeslund, Thomas B.

AU - Trivedi, Mohan M.

PY - 2015

Y1 - 2015

N2 - Traffic light recognition (TLR) is an integral part of any intelligent vehicle, which must function in the existing infrastructure. Pedestrian and sign detection have recently seen great improvements due to the introduction of learning based detectors using integral channel features. A similar push have not yet been seen for the detection sub-problem of TLR, where detection is dominated by methods based on heuristic models.Evaluation of existing systems is currently limited primarily to small local datasets. In order to provide a common basis for comparing future TLR research an extensive public database is collected based on footage from US roads. The database consists of both test and training data, totaling 46,418 frames and 112,971 annotated traffic lights, captured in continuous sequences under a varying light and weather conditions.The learning based detector achieves an AUC of 0.4 and 0.32 for day sequence 1 and 2, respectively, which is more than an order of magnitude better than the two heuristic model-based detectors.

AB - Traffic light recognition (TLR) is an integral part of any intelligent vehicle, which must function in the existing infrastructure. Pedestrian and sign detection have recently seen great improvements due to the introduction of learning based detectors using integral channel features. A similar push have not yet been seen for the detection sub-problem of TLR, where detection is dominated by methods based on heuristic models.Evaluation of existing systems is currently limited primarily to small local datasets. In order to provide a common basis for comparing future TLR research an extensive public database is collected based on footage from US roads. The database consists of both test and training data, totaling 46,418 frames and 112,971 annotated traffic lights, captured in continuous sequences under a varying light and weather conditions.The learning based detector achieves an AUC of 0.4 and 0.32 for day sequence 1 and 2, respectively, which is more than an order of magnitude better than the two heuristic model-based detectors.

U2 - 10.1109/ITSC.2015.378

DO - 10.1109/ITSC.2015.378

M3 - Article in proceeding

SN - 978-1-4673-6595-6

T3 - I E E E Intelligent Transportation Systems Conference. Proceedings

SP - 2341

EP - 2345

BT - 2015 IEEE 18th International Conference on Intelligent Transportation Systems (ITSC 2015)

PB - IEEE

ER -

Philipsen MP, Jensen MB, Møgelmose A, Moeslund TB, Trivedi MM. Traffic Light Detection: A Learning Algorithm and Evaluations on Challenging Dataset. I 2015 IEEE 18th International Conference on Intelligent Transportation Systems (ITSC 2015): Proceedings. IEEE. 2015. s. 2341-2345. 7313470. (I E E E Intelligent Transportation Systems Conference. Proceedings). https://doi.org/10.1109/ITSC.2015.378