MIoU，Mean IoU，Mean Intersection over Union，均交并比

JNingWei

发布于 2021-12-06 21:32:12

9130

发布于 2021-12-06 21:32:12

MIoU（Mean IoU，Mean Intersection over Union，均交并比，交集 / 并集），也就是语义分割中所谓的 Mask IoU 。
MIoU：计算两圆交集（橙色TP）与两圆并集（红色FN+橙色TP+黄色FP）之间的比例，理想情况下两圆重合，比例为1。
在这里插入图片描述

 from sklearn.metrics import confusion_matrix  
 import numpy as np

 def compute_iou(y_pred, y_true):
     # ytrue, ypred is a flatten vector
     y_pred = y_pred.flatten()
     y_true = y_true.flatten()
     current = confusion_matrix(y_true, y_pred, labels=[0, 1])
     # compute mean iou
     intersection = np.diag(current)
     ground_truth_set = current.sum(axis=1)
     predicted_set = current.sum(axis=0)
     union = ground_truth_set + predicted_set - intersection
     IoU = intersection / union.astype(np.float32)
     return np.mean(IoU)

本文参与腾讯云自媒体同步曝光计划，分享自作者个人站点/博客。

原始发表：2021-09-06 ，如有侵权请联系 cloudcommunity@tencent.com 删除

intersection