WEBVTT

00:00.120 --> 00:02.640
Hello developers, and welcome back.

00:02.960 --> 00:11.720
In this section we're going to build a very complex application called Custom Object Detection with

00:11.840 --> 00:12.520
YOLO.

00:12.560 --> 00:15.840
We're going to introduce YOLO to detect batteries.

00:15.840 --> 00:19.600
We're going to train our model using Google Colab.

00:19.760 --> 00:24.720
And we're going to download that data set from Roboflow.

00:24.760 --> 00:27.240
We're going to introduce the Roboflow.

00:27.280 --> 00:38.200
We're going to train our model, upload the images of our um of our object and label it, and then export

00:38.200 --> 00:41.840
this, uh, this into our model.

00:42.040 --> 00:53.000
And then using this data set with Google Colab in order to train our model and export our model as Tflite,

00:53.040 --> 01:01.960
then we're going to move into the Android Studio and import our model and the required dependencies.

01:02.160 --> 01:09.000
And it's this application is similar to what we've done before in the previous section.

01:09.000 --> 01:15.440
So we're going to, uh, see the differences between the SSD mobile.

01:15.440 --> 01:19.480
Net model and YOLO detection model.

01:19.520 --> 01:20.000
Okay.

01:20.160 --> 01:23.760
So we're going to do the updates for the codes.

01:23.760 --> 01:31.500
We're going to make some difference and comparison between the both models.

01:31.500 --> 01:38.660
We're going to learn more about YOLO object detection and we're going to introduce ultra Lytics.

01:38.860 --> 01:48.900
The main idea behind this application is to pay attention for the model any model input and output shapes.

01:48.900 --> 01:58.460
So we're going to understand the YOLO shapes, the input which is Float32 one 640 640 and three, and

01:58.460 --> 02:03.060
the output is one five 8400.

02:03.140 --> 02:09.900
And by the way, since we are only detecting one class, this is number five.

02:09.940 --> 02:13.540
If you have more and more this number will be changed.

02:13.580 --> 02:14.060
Okay.

02:14.220 --> 02:22.100
So I want from you to focus with me and understand the main idea behind this application, which is

02:22.460 --> 02:29.340
understanding and mastering input and output shapes for every model you get.

02:29.540 --> 02:33.580
Okay, so this is the main point behind this application.

02:33.580 --> 02:43.340
I promise you it will be a very, very interesting application that will increase your coding skills

02:43.340 --> 02:46.020
and machine learning concepts.

02:46.060 --> 02:48.780
Let's start building this amazing app.