You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, as part of a student games dev project, I'm looking for a linux solution which will remove everything but a face from the image, leaving only the face, and transparent alpha channel ideally.
Can backscrub extract just faces?
The virtual cam output can then be read within a browser 2D platform game for further processing (using players face on their actual player character) .
The text was updated successfully, but these errors were encountered:
Interesting question - this won't work out-of-the-box, because all the ML models are trained to segment persons and not faces. But you could try a two-step approach, where you take the person outline from backscrub, and then run a dedicated face detector such as described in https://docs.opencv.org/4.2.0/db/d28/tutorial_cascade_classifier.html (and then just take the intersection of the two regions).
Also if you take a compatible model that only marks faces in the image this might even work with backscrub directly. But depending on the model you will likely need to adapt some stuff to properly integrate the model with the internal processing pipeline we use.
Hi, as part of a student games dev project, I'm looking for a linux solution which will remove everything but a face from the image, leaving only the face, and transparent alpha channel ideally.
Can backscrub extract just faces?
The virtual cam output can then be read within a browser 2D platform game for further processing (using players face on their actual player character) .
The text was updated successfully, but these errors were encountered: