WEBVTT - generated by wenglor-media

1
00:00:00.560 --> 00:00:02.640
In this video, we will see how to export

2
00:00:02.640 --> 00:00:06.000
a trained model from AI Lab back into uniVision
3

3
00:00:06.000 --> 00:00:08.080
using the AI Loop through weHub.

4
00:00:08.240 --> 00:00:10.240
We start in uniVision 3.

5
00:00:10.320 --> 00:00:13.080
Add the “Module Image AI” to your project

6
00:00:13.080 --> 00:00:15.080
and link an input image.

7
00:00:18.960 --> 00:00:22.360
Next, paste your “AI Lab Dataset ID”

8
00:00:22.360 --> 00:00:23.960
into the corresponding field —

9
00:00:24.360 --> 00:00:26.280
this allows the module to identify

10
00:00:26.280 --> 00:00:28.840
your dataset directly from AI Lab.

11
00:00:28.920 --> 00:00:31.240
Now, you can choose which model to load —

12
00:00:31.240 --> 00:00:33.960
either the latest model or your primary model.

13
00:00:33.960 --> 00:00:36.160
In our case, since we only have one,

14
00:00:36.160 --> 00:00:37.960
it doesn’t matter which we select.

15
00:00:38.080 --> 00:00:39.480
After a short moment,

16
00:00:39.480 --> 00:00:42.560
the model is loaded into your uniVision job.

17
00:00:50.760 --> 00:00:52.120
You will see a quick preview

18
00:00:52.120 --> 00:00:54.200
of the image along with some information —

19
00:00:54.480 --> 00:00:56.840
such as class names or resolution.

20
00:00:57.360 --> 00:00:58.440
Close the dialog

21
00:00:58.440 --> 00:01:01.000
and the image will be classified instantly.

22
00:01:01.000 --> 00:01:04.000
Click on “Switch to Run Mode” to see the
live results

23
00:01:04.000 --> 00:01:06.000
and watch as the system distinguishes

24
00:01:06.000 --> 00:01:10.000
between your OK and NOK classes in real time.

25
00:01:10.000 --> 00:01:11.000
That’s it —

26
00:01:11.000 --> 00:01:13.000
your AI Lab model is now fully deployed

27
00:01:13.000 --> 00:01:15.000
and running on uniVision 3!