IIntersection over Union (IoU) is a metric used to quantify the accuracy of an object detector on a particular example by comparing the predicted bounding box with the ground truth bounding box. It’s defined as the area of the overlap between the predicted box and the ground truth box divided by the area of their union. Mathematically, IoU = (Area of Overlap) / (Area of Union). IoU is 1 if boxes perfectly coincide, and 0 if they don’t overlap at all. In evaluation, a prediction is considered a “true positive” if IoU exceeds a certain threshold (commonly 0.5) with a ground truth box of the same class. IoU is also used in segmentation to compare predicted vs GT masks (also known as Jaccard Index). It provides a measure of how well an object is localized and is robust to boxes being slightly off as long as the significant majority overlaps. Higher IoU thresholds demand stricter localization. Average Precision (AP) in detection is often computed by considering IoU ≥ 0.5 as positive, and mAP can consider multiple IoU thresholds.
Data Selection & Data Viewer
Get data insights and find the perfect selection strategy
Learn MoreSelf-Supervised Pretraining
Leverage self-supervised learning to pretrain models
Learn MoreSmart Data Capturing on Device
Find only the most valuable data directly on devide
Learn More