I Wanted to share this with you so every one who have issue with Object Detection templet can now create something.
I'm gonna talk about Object Detection using the powerful Snap ML. and we will start talking about the basics.
Before we start check out the object detection Guides :
So first we have to download the notebook so we can train the model:
In file we will see two important elements: the ( Car test image ) and the file with the ( ipynb ) extension.
The file has as an example recognizing cars image
In this case we will recognize book or what ever you would like So make sure the photo you will take must have a good light and with the objects you are going to work on in my case i use the photo with extension JPG and make sure to name it to " book_test "
So the platform where to start the training the model is " Google Colaboratory "
When we enter we can easily upload the work file " the file with the ( ipynb ) extension. "
Then automatically we will have this view the codes and their explanations for each section. that way we understand better what we are doing
On " Object labels union " we will choose a set of related objects as in this case we will add " book "
Then we go to the point 10 and change it to ( './book_test.jpg' )
Make sure to click on the file box left the website. and drag the " book " image inside
Now we go to click on Runtime then Run all. When its done look for ONNX file and download it.
Now we open Lens Studio and open Object detection template the template is already predetermined to detect cars.
The first thing we will do is click on ML Controller
and here we will use our previous created model and drag it on model.
Now we are going to hide other graphic elements such as loader, hint and other components.
We will leave the " detection box " in effect and change the icon to book icon as we work on.
That's it!! I hope this helped ! :)
Open Project file :
Please sign in to leave a comment.