Original subtitles

Hello and welcome to this video.

We will explain.

YOLO v seven Training on Custom Objects.

The training will be done on windows at this time.

Make sure all of seven is installed on windows.

However, before we begin training we must first prepare the annotated dataset.

The face mask dataset was used in this example.

The dataset can be accessed and downloaded at the following URL.

Download the following dataset.

Wait until the download is finished.

When you're finished, go to the downloads folder.

This is a face mask deficit that has been annotated.

Next extract the dataset.

The dataset will be saved in the data folder of YOLO 5/7 root folder.

In this example on the YOLO five seven Dpu data, we will use tools on Windows 11 for extraction to

extract right click and select.

Extract or.

Click browse.

On the.

You know, he's 74.

You.

Data.

Click Select folder.

Click extra.

Wait until the extraction is finished.

When finished, it will be saved in the data folder.

The following is the annotated face mask dataset.

Following the split the dataset into train validation and test data.

The split results must match the goal of seven folder structure.

The all of seven folder structure is shown below.

The images folder contains images and the labels folder contains annotations.

For splitting in its folder.

There are three more folders.

Specifically the train well and test follows.

We have provided Python code to split the dataset.

Specifically split dataset dot pie.

In this example, we open the code with Fisher's studio code, you can also use another text editor.

Here is the code.

There are several arguments that can be used.

The first argument is the train argument, which is used to specify the percentage of train data.

The default value is 80.

Then there is the validation argument, which is used to specify the percentage of data validation.

The default value is ten.

Then there is the test argument, which is used to specify the percentage of data tests.

The default value is ten.

The further argument is used to specify the folder where the dataset is stored before splitting.

The This argument is used to specify the folder where the split results will be safe.

After that, we try to do the splitting with the code.

Press the windows key, then type in a condom.

Click on the Anaconda prompt.

After that activate the all of seven CPU environment using the command.

Activate.

YOLO v seven to for you and v.

Press enter.

Then navigate to the data folder.

We will split the death of it.

Composition 80%.

Train death 10% validation data and 10% test data using the command.

Clayton Split Data set dot p.

In the further argument, we read the face mask.

In the train argument we wrote at the.

Invalidates an argument.

We write ten.

In the test argument we write ten.

We will save the split result in the Face mask dataset folder.

First intro.

Then we try to see the results.

Here is the Face Mask dataset folder.

There is an images and labels folder.

And there are three speed folders.

The following is an example of the photo's contents.

In the next video, we will create a configuration file.

See you then.