Closed shukkkur closed 1 year ago
Hi @shukkkur, Present you can figure it in c++ version as mentioned here https://github.com/google/mediapipe/issues/2028#issuecomment-841435348.
Hello @sgowroji Can I also remove hands and track for example just legs and torso? Where exactly shall I uncomment mentioned code? Thanks
Hey how do I work on this?
Bump. @shukkkur did you ever manage to do this? We use MediaPipe with Unity, but are also wondering if we can somehow remove face from detection. We are experiences weird issues. If someone is standing still, and rotates only their head to the side, coordinates for the shoulders(which we use to detect the angle) also change. We are hoping that by removing face from detection this should perhaps be better.
@patrick508 Hey there! Nah, unfortunately I couldn't find anything.
I am still sure that the idea of removing head from detection should improve the accuracy overall.
In case you find anything, keep us posted!
😄
Ah to bad! I will definitely keep you posted if I find anything! Thanks for the reply!
@shukkkur @patrick508 I was thinking that removing the face might improve performance.
Hello @shukkkur, We are upgrading the MediaPipe Legacy Solutions to new MediaPipe solutions However, the libraries, documentation, and source code for all the MediaPipe 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.
The issue is not resolved and quite unfortunate that none of the developers got to see it before it is closed. Now definitely no one is gonna solve it.
System information
Currently if the mediapipe pose estimation can't detect the face, the whole detection fails. It would be nice to have a detection model that tries to detect only the body.
use cases for this feature: I am currently working on a workout-counter project, so whenever the barbell covers the face the detection fails and it least to more false negatives result (counter doesn't count the rep) Similar problem is with the bench press