Integrate Autotags assignment feature into Maintenance testing tools and integration testing to benchmark the Deep-Learning version (currently Yolo Versions)

Hi all,

I intend to create the MR (Merge Request) for the integration testing tools to benchmark the YOLO version. The tools will include:

• The control panel with options to choose the model and 80 predefined classes.
• An image area to display all the detected objects. (Any suggested settings to add to the widget would be appreciated.)

The performance of the tools is better than what I researched in Python in terms of inference time per image. Currently, there are two models with the following performance on CPU, on average:

• YOLOv5 Nano: 57 ms
• YOLOv5 XLarge: 724 ms

In my opinion, YOLOv5 XLarge is very good for detect the complex image, the users who want to detect in a few of image, that release the appropriate answers.

Here is a video capture I would like to show you:

test

Updatae

• Integrated autotagging features into Maintenance tools.

Coding file structures:

• Auto tags Back-end logic located in core/libs/autotagsassignment
• cli an gui test located in core/test/autotagsassignment
• Intergrate in Maintenance tools located in core/utilities/maintenance