Keywords: Path Tracing, Differentiable Path Tracing, Gaussian Smoothing, Variational Optimization, Differentiable Light Transport, Rendering, Differentiable Rendering
TL;DR: We convolve the traditional rendering equation with a Gaussian kernel, which smoothens the light-path space and reduce plateaus in the objective function of differentiable inverse rendering, improving convergence.
Abstract: Current differentiable renderers provide light transport gradients with respect to arbitrary scene parameters. However, the mere existence of these gradients does not guarantee useful update steps in an optimization. Instead, inverse rendering might not converge due to inherent plateaus, i.e., regions of zero gradient, in the objective function. We propose to alleviate this by convolving the high-dimensional rendering function, that maps scene parameters to images, with an additional kernel that blurs the parameter space. We describe two Monte Carlo estimators to compute plateau-reduced gradients efficiently, i.e., with low variance, and show that these translate into net-gains in optimization error and runtime performance. Our approach is a straightforward extension to both black-box and differentiable renderers and enables optimization of problems with intricate light transport, such as caustics or global illumination, that existing differentiable renderers do not converge on.
Submission Number: 63
Loading