Estimating and Exploiting the Aleatoric Uncertainty in Surface Normal Estimation

TL;DR

This paper introduces a novel method for surface normal estimation from a single image that estimates aleatoric uncertainty and uses it to improve prediction quality, especially around boundaries and small structures.

Contribution

It proposes a new distribution parameterization for uncertainty estimation and a decoder framework that leverages this uncertainty for better surface normal predictions.

Findings

Outperforms state-of-the-art on ScanNet and NYUv2 datasets.

Estimated uncertainty correlates well with actual prediction error.

Improves detail and accuracy near object boundaries.

Abstract

Surface normal estimation from a single image is an important task in 3D scene understanding. In this paper, we address two limitations shared by the existing methods: the inability to estimate the aleatoric uncertainty and lack of detail in the prediction. The proposed network estimates the per-pixel surface normal probability distribution. We introduce a new parameterization for the distribution, such that its negative log-likelihood is the angular loss with learned attenuation. The expected value of the angular error is then used as a measure of the aleatoric uncertainty. We also present a novel decoder framework where pixel-wise multi-layer perceptrons are trained on a subset of pixels sampled based on the estimated uncertainty. The proposed uncertainty-guided sampling prevents the bias in training towards large planar surfaces and improves the quality of prediction, especially near object boundaries and on small structures. Experimental results show that the proposed method outperforms the state-of-the-art in ScanNet and NYUv2, and that the estimated uncertainty correlates well with the prediction error. Code is available at https://github.com/baegwangbin/surface_normal_uncertainty.