Face Landmark Detection & Pose Estimation With Raspberry Pi + OpenCV + Python

Can we use this to detect where a person is looking at ( gaze tracking)

I want to make a bodyguard robot that protects me by putting punch and aggression detection
when the robot detects a person approaching me to hit me
he tries to shoot him with airsoft balls
so is there punch detection, we will say that I am a victim and a person approaches and punches me
does the camera detect that it is a punch and that the robot must react by shooting it with airsoft balls and still in the camera I am the victim even if I hit it the robot does not shoot me not on it and he always remains focused on the person who attacks me by shooting him until he leaves
and how i code all this on the raspberry pi
please if you can help me

Great!

Great video!
Is it possible to recognize a logo on a golf ball, so I can send a signal to a motor and consequently have the logo on the place I want?
Thanks in advance

Awesome video but the link to the Full Article points to another article: "Hand Recognition and Finger Identification with Raspberry Pi and OpenCV" and not to "Face Landmark Detection & Pose Estimation With Raspberry Pi + OpenCV". In addition to that, I can't find where to search for articles on your site

Does adding nodes for facial detection decrease the speed in which the program runs?

Amazing quality content!

Can we do face recognition on rpi4 from 2 cameras?
And will the performance depend upon the ram of the board

Would you be able to show us how to do this using opencv on the 'Bullseye' OS?

How much different is the performance if I only have a Raspberry Pi 3b+

I cant' thank you enough. You managed to resolve many problems I had doing something similar with a raspberry Pi. Thank you.

very cool

Great post.

Genial 👌🏻

Possible application for mute people, hand-sign to text or speech.