BOUNDING BOX DISPARITY: 3D METRICS FOR OBJECT DETECTION WITH FULL DEGREE OF FREEDOM

Michael G. Adam, Martin Piccolrovazzi, Sebastian Eger, Eckehard Steinbach

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

2 Zitate (Scopus)

Abstract

The most popular evaluation metric for object detection in 2D images is Intersection over Union (IoU). Existing implementations of the IoU metric for 3D object detection usually neglect one or more degrees of freedom. In this paper, we first derive the analytic solution for three dimensional bounding boxes. As a second contribution, a closed-form solution of the volume-to-volume distance is derived. Finally, the Bounding Box Disparity is proposed as a combined positive continuous metric. We provide open source implementations of the three metrics as standalone python functions, as well as extensions to the Open3D library and as ROS nodes.

OriginalspracheEnglisch
Titel2022 IEEE International Conference on Image Processing, ICIP 2022 - Proceedings
Herausgeber (Verlag)IEEE Computer Society
Seiten1491-1495
Seitenumfang5
ISBN (elektronisch)9781665496209
DOIs
PublikationsstatusVeröffentlicht - 2022
Veranstaltung29th IEEE International Conference on Image Processing, ICIP 2022 - Bordeaux, Frankreich
Dauer: 16 Okt. 202219 Okt. 2022

Publikationsreihe

NameProceedings - International Conference on Image Processing, ICIP
ISSN (Print)1522-4880

Konferenz

Konferenz29th IEEE International Conference on Image Processing, ICIP 2022
Land/GebietFrankreich
OrtBordeaux
Zeitraum16/10/2219/10/22

Fingerprint

Untersuchen Sie die Forschungsthemen von „BOUNDING BOX DISPARITY: 3D METRICS FOR OBJECT DETECTION WITH FULL DEGREE OF FREEDOM“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren