Bounding Box Disparity: 3D Metrics for Object Detection With Full Degree of Freedom
This work provides improved evaluation tools for 3D object detection, which is incremental as it builds on existing IoU metrics by adding full degree-of-freedom support.
The paper tackles the problem of evaluating 3D object detection by addressing limitations in existing Intersection over Union (IoU) metrics, which often neglect degrees of freedom, and proposes new metrics including an analytic solution for 3D bounding boxes, a closed-form volume-to-volume distance, and the Bounding Box Disparity as a combined positive continuous metric.
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.