mediapipe

mediapipe copied to clipboard

Hello! As I understand from documentation, AttentionModule produces 478 landmars face model, adding Irises to the standard one. mp_face_mesh.FACEMESH_NUM_LANDMARKS = 468 mp_face_mesh.FACEMESH_NUM_LANDMARKS_WITH_IRISES = 478 However, mp_face_mesh.FACEMESH_IRISES=8.

My question is what are those two missing landmarks? I have 468 landmarks canonical face model and I want to solve PnP. I could not find 478 landmarks canonical face model. If You can provide me with such one, it will be great!

Best, Jan

Hi @janglinko-dac, Could you please elaborate your query with complete details. Thank you!

Hi @kuaashish, thanks for the response. I want to estimate a head pose in the camera frame. I want to use FaceMesh module to detect landmarks and then, with use of a canonical facel model, solve PnP to get the position.

In the documentation is it said:

In addition to the Face Landmark Model we provide another model that applies attention to semantically meaningful face regions, and therefore predicting landmarks more accurately around lips, eyes and irises, at the expense of more compute. It enables applications like AR makeup and AR puppeteering.

The attention mesh model can be selected in the Solution APIs via the refine_landmarks option.

Attention mesh results in face 478 landmarsk. mp_face_mesh.FACEMESH_NUM_LANDMARKS_WITH_IRISES = 478

My canonical face model has 468 landmarks (like the output with refine_landmarks=False), but i want to use attention mesh, as results are more accurate.

I would like to remove landmarks added by attention mesh, but there are only 8 Iris landmarks. So, 2 landmarks left to remove and i don't know their numbers.

My question is about the numbers of these 2 landmarks.

Hello @janglinko-dac, We are upgrading the MediaPipe Legacy Solutions to new MediaPipe solutions However, the libraries, documentation, and source code for all the MediapPipe Legacy Solutions will continue to be available in our GitHub repository and through library distribution services, such as Maven and NPM.

You can continue to use those legacy solutions in your applications if you choose. Though, we would request you to check new MediaPipe solutions which can help you more easily build and customize ML solutions for your applications. These new solutions will provide a superset of capabilities available in the legacy solutions. Thank you

This issue has been marked stale because it has no recent activity since 7 days. It will be closed if no further activity occurs. Thank you.

This issue was closed due to lack of activity after being marked stale for past 7 days.

Are you satisfied with the resolution of your issue? Yes No