1
00:00:01,000 --> 00:00:06,000
This lecture presents an introduction to object segmentation using YOLO V8.

2
00:00:06,000 --> 00:00:09,000
Let's first look at the objectives of this lecture.

3
00:00:12,000 --> 00:00:14,000
We have divided this complete lecture into three parts.

4
00:00:14,000 --> 00:00:21,000
In the first part, we will see how we can implement object segmentation using YOLO V8 in Google CoLab.

5
00:00:21,000 --> 00:00:27,000
In the second part, we will see what are the different steps involved in training the YOLO V8 segmentation

6
00:00:27,000 --> 00:00:29,000
model in your own custom dataset.

7
00:00:30,000 --> 00:00:35,000
We will train the YOLO V8 segmentation model on potholes dataset in this section.

8
00:00:35,000 --> 00:00:42,000
In the third part I will play the output video of the trained YOLO V8 segmentation model testing on

9
00:00:42,000 --> 00:00:43,000
a video.

10
00:00:43,000 --> 00:00:48,000
The YOLO V8 segmentation model is trained on the potholes dataset.

11
00:00:49,000 --> 00:00:53,000
So let's look how to implement object segmentation using YOLO V8.

12
00:00:53,000 --> 00:00:59,000
The YOLO V8 segmentation model is one of the fastest and most accurate models for real time instance

13
00:00:59,000 --> 00:01:00,000
segmentation.

14
00:01:01,000 --> 00:01:07,000
Segmentation goes a step further than object detection and involves identifying individual objects in

15
00:01:07,000 --> 00:01:11,000
an image and segmenting them from the rest of the image.

16
00:01:11,000 --> 00:01:17,000
And in the below you can see the three images which differentiates what is basically classification.

17
00:01:17,000 --> 00:01:18,000
The other is object detection.

18
00:01:18,000 --> 00:01:21,000
And the third one is segmentation.

19
00:01:21,000 --> 00:01:26,000
So if you see these three images, you can know the difference between classification object detection

20
00:01:26,000 --> 00:01:28,000
and the segmentation.

21
00:01:29,000 --> 00:01:32,000
Implement segmentation using YOLO V8.

22
00:01:32,000 --> 00:01:38,000
Just write this two lines of code which you are seeing in front of your screen on any of the ID and

23
00:01:38,000 --> 00:01:45,000
pass input image or the video and the segmentation on the input image and video will be done.

24
00:01:47,000 --> 00:01:52,000
To implement object segmentation using V8 on custom dataset.

25
00:01:52,000 --> 00:01:54,000
Here are the different steps involved.

26
00:01:54,000 --> 00:01:56,000
In the first step, we prepare the dataset.

27
00:01:56,000 --> 00:02:03,000
In the second step, we fine tune the YOLO V8 segmentation on module V8 segmentation model on custom

28
00:02:03,000 --> 00:02:06,000
dataset, so just fine tuning means training.

29
00:02:06,000 --> 00:02:11,000
The YOLO segmentation model on the custom dataset which we have prepared.

30
00:02:11,000 --> 00:02:17,000
In the third step, we validate the fine tuned YOLO V8 model and see what is the mean average precision

31
00:02:17,000 --> 00:02:22,000
we got and how our model performs on random images and videos.

32
00:02:22,000 --> 00:02:29,000
In the fourth step, we test the model on some sample images and videos and display those images and

33
00:02:29,000 --> 00:02:31,000
videos as our outputs.

34
00:02:32,000 --> 00:02:38,000
So here is the output video of the train segmentation model testing on a demo video.

35
00:02:38,000 --> 00:02:44,000
That YOLO V8 segmentation model is trained on the potholes dataset and you can see that the model is

36
00:02:44,000 --> 00:02:48,000
able to detect and do the segmentation of the potholes.

37
00:02:49,000 --> 00:02:51,000
Thank you so much for watching this video tutorial.

38
00:02:51,000 --> 00:02:53,000
See you all in the next video tutorial.

39
00:02:53,000 --> 00:02:55,000
Till then, bye bye.