Abstract
In the 2D/3D object detection task, Intersection-over-Union (IoU) has been widely employed as an evaluation metric to evaluate the performance of different detectors in the testing stage. However, during the training stage, the common distance loss (e.g, L-1 or L-2) is often adopted as the loss function to minimize the discrepancy between the predicted and ground truth Bounding Box (Bbox). To eliminate the performance gap between training and testing, the IoU loss has been introduced for 2D object detection in [1] and [2]. Unfortunately, all these approaches only work for axis-aligned 2D Boxes, which cannot be applied for more general object detection task with rotated Boxes. To resolve this issue, we investigate the IoU computation for two rotated Boxes first and then implement a unified framework, IoU loss layer for both 2D and 3D object detection tasks. By integrating the implemented IoU loss into several state-of-the-art 3D object detectors, consistent improvements have been achieved for both bird-eye-view 2D detection and point cloud 3D detection on the public KITTI [3] benchmark.
Original language | English |
---|---|
Title of host publication | Proceedings - 2019 International Conference on 3D Vision, 3DV 2019 |
Pages | 85-94 |
Number of pages | 10 |
ISBN (Electronic) | 9781728131313 |
DOIs | |
State | Published - Sep 2019 |
Event | 7th International Conference on 3D Vision, 3DV 2019 - Quebec, Canada Duration: Sep 15 2019 → Sep 18 2019 |
Publication series
Name | Proceedings - 2019 International Conference on 3D Vision, 3DV 2019 |
---|
Conference
Conference | 7th International Conference on 3D Vision, 3DV 2019 |
---|---|
Country/Territory | Canada |
City | Quebec |
Period | 9/15/19 → 9/18/19 |
Bibliographical note
Publisher Copyright:© 2019 IEEE.
Keywords
- 3D Object Detection
- Autonomous Driving
- IoU Loss
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition
- Media Technology
- Modeling and Simulation