search

Qengineering / Face-Recognition-Jetson-Nano

Recognize 2000+ faces on your Jetson Nano with database auto-fill and anti-spoofing
https://qengineering.eu/deep-learning-examples-on-raspberry-32-64-os.html
BSD 3-Clause "New" or "Revised" License
108 stars 31 forks source link

Add face tracking to application #11

Open fadsamo opened 2 years ago

fadsamo commented 2 years ago

Hi First of all, thank you for this great work I want to track every face that is detected and save only one face of a particular id. How can I do this?

Qengineering commented 2 years ago

Dear @fadsamo ,

Can you explain me in more detail what precisely you want?

The app stores all the faces in a 'database' located at ./img/ folder. You may manipulate the database by hand, choosing the faces you want to be recognized. But, I presume that's not want you intended to do.

fadsamo commented 2 years ago

I do not want to apply the face recognition module. I just want to track the faces that can be detected by retinaface or Mycenaean module and save the best face that is not side (frontal), or not blurry , or has more details ... But the problem is that for each face in front of the camera, I save only one or two faces for each ID and finally send these faces detected by jetson nano to the cloud to perform face recognition ... I know that using kalman tracking algorithms, you can track objects, and here these objects can be face landmarks. But I do not know how to do this in your program.

Qengineering commented 2 years ago

Dear @fadsamo, Now it's clear to me want you trying to build. However, it is not something you can fix in a few lines of code. In fact, it is far beyond the scope of this project and could be a complete new project. My apologies, I can't help you on this one. Try to find another repo on GitHub with tracking already available, like people counters.