Self-Calibrating, Fully Differentiable NLOS Inverse Rendering

TL;DR

This paper presents a fully differentiable NLOS inverse rendering pipeline that self-calibrates imaging parameters, enabling robust reconstruction of hidden scene geometry and albedo from indirect illumination measurements, even with noise.

Contribution

It introduces a novel end-to-end differentiable framework that jointly reconstructs geometry and estimates imaging parameters without manual tuning.

Findings

Robust reconstruction of hidden scene geometry and albedo under noise.

End-to-end pipeline effectively self-calibrates imaging parameters.

Accurate dense surface point and normal extraction from NLOS data.

Abstract

Existing time-resolved non-line-of-sight (NLOS) imaging methods reconstruct hidden scenes by inverting the optical paths of indirect illumination measured at visible relay surfaces. These methods are prone to reconstruction artifacts due to inversion ambiguities and capture noise, which are typically mitigated through the manual selection of filtering functions and parameters. We introduce a fully-differentiable end-to-end NLOS inverse rendering pipeline that self-calibrates the imaging parameters during the reconstruction of hidden scenes, using as input only the measured illumination while working both in the time and frequency domains. Our pipeline extracts a geometric representation of the hidden scene from NLOS volumetric intensities and estimates the time-resolved illumination at the relay wall produced by such geometric information using differentiable transient rendering. We then use gradient descent to optimize imaging parameters by minimizing the error between our simulated time-resolved illumination and the measured illumination. Our end-to-end differentiable pipeline couples diffraction-based volumetric NLOS reconstruction with path-space light transport and a simple ray marching technique to extract detailed, dense sets of surface points and normals of hidden scenes. We demonstrate the robustness of our method to consistently reconstruct geometry and albedo, even under significant noise levels.