Flow supervision for Deformable NeRF
This work addresses the problem of computationally inefficient flow supervision in deformable NeRF for researchers in 3D reconstruction and view synthesis, representing an incremental improvement by simplifying existing methods.
The paper tackles the challenge of efficiently using optical flow as supervision for deformable NeRF by eliminating the need to invert backward deformation functions, enabling application to a broad class of deformation fields. It demonstrates significant improvements in monocular novel view synthesis with rapid object motion over baselines without flow supervision.
In this paper we present a new method for deformable NeRF that can directly use optical flow as supervision. We overcome the major challenge with respect to the computationally inefficiency of enforcing the flow constraints to the backward deformation field, used by deformable NeRFs. Specifically, we show that inverting the backward deformation function is actually not needed for computing scene flows between frames. This insight dramatically simplifies the problem, as one is no longer constrained to deformation functions that can be analytically inverted. Instead, thanks to the weak assumptions required by our derivation based on the inverse function theorem, our approach can be extended to a broad class of commonly used backward deformation field. We present results on monocular novel view synthesis with rapid object motion, and demonstrate significant improvements over baselines without flow supervision.