Real time image manipulation using OpenCV

JordiCP

Expert
Licensed User
Longtime User
Perhaps our OpenCV expert @JordiCP would like to comment on it also? :)
No, thanks ?
(joking)

The fact is that there are (at least that I know of) a couple of members in this forum who are more experts than me in OpenCV. I just made the wrapper and some interesting projects with it, but still exploring some modules (btw, I'm preparing a project that I'm sure you will like, but is going slow, stay tuned ?)

Seems that it is using some keypoints (face landmarks) from the video image, also from the target image, and compute the time variations of the former to somehow translate them to the target image and "gracefully" apply transformations based on these. The way it is done -how it detects them, using ML models, DNN...- and how they are projected into the target image, and I guess that many other things, are 'the sauce' of it.

Really impressed, indeed ?
 
Top