MinBackProp -- Backpropagating through Minimal Solvers

arXiv:2404.17993v1 Announce Type: new
Abstract: We present an approach to backpropagating through minimal problem solvers in end-to-end neural network training. Traditional methods relying on manually constructed formulas, finite differences, and autograd are laborious, approximate, and unstable for complex minimal problem solvers. We show that using the Implicit function theorem to calculate derivatives to backpropagate through the solution of a minimal problem solver is simple, fast, and stable. We compare our approach to (i) using the standard autograd on minimal problem …

abstract arxiv autograd cs.cv derivatives differences function network network training neural network show theorem through training type