Abstract: Computing the matrix square root and its inverse in a differentiable manner is important in a variety of computer vision tasks. Previous methods either adopt the Singular Value Decomposition ...
Results that may be inaccessible to you are currently showing.
Hide inaccessible results
Feedback