Multi-Perspective Vehicle Detection and Tracking: Challenges, Dataset, and Metrics

Jacob Velling Dueholm, Miklas Strøm Kristoffersen, Ravi K. Satzoda, Eshed Ohn-Bar, Thomas B. Moeslund, Mohan M. Trivedi

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

8 Citationer (Scopus)


The research community has shown significant improvements
in both vision-based detection and tracking of vehicles,
working towards a high level understanding of on-road
maneuvers. Behaviors of surrounding vehicles in a highway
environment is found as an interesting starting point, of
why this dataset is introduced along with its challenges
and evaluation metrics. A vision-based multi-perspective
dataset is presented, containing a full panoramic view from
a moving platform driving on U.S. highways capturing
2704x1440 resolution images at 12 frames per second. The
dataset serves multiple purposes to be used as traditional
detection and tracking, together with tracking of vehicles
across perspectives. Each of the four perspectives have
been annotated, resulting in more than 4000 bounding boxes
in order to evaluate and compare novel methods.
Titel2016 IEEE 19th International Conference on Intelligent Transportation Systems
Publikationsdato26 dec. 2016
ISBN (Elektronisk)978-1-5090-1889-5
StatusUdgivet - 26 dec. 2016
BegivenhedInternational Conference on Intelligent Transportation Systems - Convention Center, Windsor Oceanico Hotel, Rio de Janeiro, Brasilien
Varighed: 1 nov. 20164 nov. 2016
Konferencens nummer: 19


KonferenceInternational Conference on Intelligent Transportation Systems
LokationConvention Center, Windsor Oceanico Hotel
ByRio de Janeiro


Dyk ned i forskningsemnerne om 'Multi-Perspective Vehicle Detection and Tracking: Challenges, Dataset, and Metrics'. Sammen danner de et unikt fingeraftryk.