LS-Net: Learning to Solve Nonlinear Least Squares for Monocular Stereo
This work addresses optimization difficulties in computer vision for researchers and practitioners, offering a learned alternative to traditional methods, though it appears incremental as it builds on existing neural optimization approaches.
The authors tackled the challenge of optimizing sum-of-squares objective functions in computer vision, which are often difficult due to unsatisfied assumptions and ill-posed problems, by proposing LS-Net, a neural nonlinear least squares optimization algorithm that learns to optimize these cost functions without hand-crafted regularizers, and applied it to motion stereo for jointly estimating motion and scene geometry from monocular image pairs, showing efficient and effective solutions.
Sum-of-squares objective functions are very popular in computer vision algorithms. However, these objective functions are not always easy to optimize. The underlying assumptions made by solvers are often not satisfied and many problems are inherently ill-posed. In this paper, we propose LS-Net, a neural nonlinear least squares optimization algorithm which learns to effectively optimize these cost functions even in the presence of adversities. Unlike traditional approaches, the proposed solver requires no hand-crafted regularizers or priors as these are implicitly learned from the data. We apply our method to the problem of motion stereo ie. jointly estimating the motion and scene geometry from pairs of images of a monocular sequence. We show that our learned optimizer is able to efficiently and effectively solve this challenging optimization problem.