课程: AutoML: Build Production-Ready Models Quickly!
今天就学习课程吧!
今天就开通帐号,24,100 门业界名师课程任您挑!
Using your model
- [Instructor] So far, we have trained the model to classify African fashion items across eight classes. We looked at how the probability threshold affects the overall recall and precision of our model. Now, we want to explore how we can use this trained model to classify new data. We navigate to the predictions tab to do this. Let's start with a quick test for a train model. We'll click quick test and then upload an image. We'll choose an image from the validation set of our data set here. So I'll click vow and then I'll choose from any of our classes. Let's go with your blouse and I'll pick your random image and upload it. Okay, so we have a probability distribution just like we saw before and we can see that very high probability is assigned to the blouse class and this prediction is correct. Let's try one more quick test but this time with a slightly different image and also an image from the internet. I'll open up…
随堂练习,边学边练
下载课堂讲义。学练结合,紧跟进度,轻松巩固知识。
内容
-
-
-
-
-
Computer vision: How do you handle image data?4 分钟 26 秒
-
(已锁定)
Introducing Azure Custom Vision4 分钟 13 秒
-
(已锁定)
Uploading and labeling images4 分钟 29 秒
-
(已锁定)
Training and validating your model5 分钟 27 秒
-
(已锁定)
Using your model4 分钟 58 秒
-
(已锁定)
Working with code: Tradeoffs3 分钟 13 秒
-
(已锁定)
Challenge: Multi-class classification56 秒
-
(已锁定)
Solution: Multi-class classification5 分钟 2 秒
-
(已锁定)
Cleaning up your resources2 分钟 52 秒
-
-
-