如何从分割掩码中找到 IoU? [英] How to find IoU from segmentation masks?

问题描述

我正在执行图像分割任务，并且我使用的数据集只有基本事实但没有边界框或多边形.

I am doing an image segmentation task and I am using a dataset that only has ground truths but no bounding boxes or polygons.

我有 2 个类(忽略 0 作为背景)并且输出和真实标签在一个数组中，如

I have 2 classes( ignoring 0 for background) and the outputs and ground truth labels are in an array like

预测--/---标签

<代码>0|0|0|1|2 0|0|0|1|20|2|1|0|0 0|2|1|0|00|0|1|1|1 0|0|1|1|10|0|0|0|1 0|0|0|0|1

我如何根据这些计算 IoU?

How do I calculate IoU from these ?

PS:我使用 python3 和 pytorch api

PS: I am using python3 with pytorch api

推荐答案

所以我才发现 jaccard_similarity_score 被认为是 IoU.

So I just found out that jaccard_similarity_score is regarded as IoU.

所以解决方案很简单，

from sklearn.metrics import jaccard_similarity_scorejac = jaccard_similarity_score(predictions, label, Normalize = True/False)

源链接:https://scikit-learn.org/stable/modules/generated/sklearn.metrics.jaccard_score.html#sklearn.metrics.jaccard_score

Source link: https://scikit-learn.org/stable/modules/generated/sklearn.metrics.jaccard_score.html#sklearn.metrics.jaccard_score

这篇关于如何从分割掩码中找到 IoU?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！