ARCore defines only three regions that can be used to get poses. They are
But how to get other regions like eyes, center of fore head, etc?
A similar question was asked in Github. That will solve the above issue. Refer the following link