Jaccard Index / Intersection over Union

IoU is a measure of the overlap between two bounding boxes (usually between prediction and ground truth), used a lot in Object Detection.

Higher IoU indicates the predicted bounding box coordinates closely resembles the ground truth box coordinates.

$\text{IoU}=\frac{\text{Area of Intersection}}{\text{Area of Union}}$

The Jaccard index, also known as the Jaccard similarity coefficient, is a statistic used for gauging the similarity and diversity of sample sets. $J(A,B)=\frac{|A\cap B|}{|A\cup B|}$

However, they are identical in generally taking the ratio of Intersection over Union

3D IoU??

For Lidar Detection, they also need to measure IoU.

Dice Coefficient

Found from here: https://www.kaggle.com/competitions/severstal-steel-defect-detection/overview/evaluation

$\frac{2|X\cap Y|}{|X|+|Y|}$