plutoyuxie / AutoEncoder-SSIM-for-unsupervised-anomaly-detection-

Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders
123 stars 25 forks source link

请问,既然需要用mask,为什么还叫无监督呢? #6

Closed FrankFirstLearn closed 4 years ago

plutoyuxie commented 4 years ago

您好,这里的mask是后处理用到的背景蒙版,不是必须的。 训练过程只用到正样本,没有使用groundtruth label信息。 请您先看看代码。

FrankFirstLearn commented 4 years ago

那我执行 test.py的时候,为什么没有那些mask文件,就会报错呢?代码有点多,我一时看不明白,对了,请问你那个一开始那一大堆信息怎么不显示啊?谢谢😁

---原始邮件--- 发件人: "plutoyuxie"<notifications@github.com> 发送时间: 2020年8月4日(周二) 中午1:25 收件人: "plutoyuxie/AutoEncoder-SSIM-for-unsupervised-anomaly-detection-"<AutoEncoder-SSIM-for-unsupervised-anomaly-detection-@noreply.github.com>; 抄送: "FrankFirstLearn"<2414345823@qq.com>;"Author"<author@noreply.github.com>; 主题: Re: [plutoyuxie/AutoEncoder-SSIM-for-unsupervised-anomaly-detection-] 请问,既然需要用mask,为什么还叫无监督呢? (#6)

您好,这里的mask是后处理用到的背景蒙版,不是必须的。 训练过程只用到正样本,没有使用groundtruth label信息。 请您先看看代码。

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.