Uncertainty-Aware CNNs for Depth Completion: Uncertainty from Beginning to End

TL;DR

This paper introduces a novel uncertainty-aware depth completion method that models uncertainty from noisy input to final prediction, outperforming existing Bayesian approaches in accuracy, uncertainty quality, and efficiency.

Contribution

It proposes a self-supervised confidence estimator and a probabilistic NCNN, enabling accurate uncertainty quantification and improved depth completion performance.

Findings

Outperforms existing Bayesian methods in accuracy and uncertainty quality

Achieves state-of-the-art results with a small network of 670k parameters

Provides efficient and reliable depth predictions with meaningful uncertainty estimates

Abstract

The focus in deep learning research has been mostly to push the limits of prediction accuracy. However, this was often achieved at the cost of increased complexity, raising concerns about the interpretability and the reliability of deep networks. Recently, an increasing attention has been given to untangling the complexity of deep networks and quantifying their uncertainty for different computer vision tasks. Differently, the task of depth completion has not received enough attention despite the inherent noisy nature of depth sensors. In this work, we thus focus on modeling the uncertainty of depth data in depth completion starting from the sparse noisy input all the way to the final prediction. We propose a novel approach to identify disturbed measurements in the input by learning an input confidence estimator in a self-supervised manner based on the normalized convolutional neural networks (NCNNs). Further, we propose a probabilistic version of NCNNs that produces a statistically meaningful uncertainty measure for the final prediction. When we evaluate our approach on the KITTI dataset for depth completion, we outperform all the existing Bayesian Deep Learning approaches in terms of prediction accuracy, quality of the uncertainty measure, and the computational efficiency. Moreover, our small network with 670k parameters performs on-par with conventional approaches with millions of parameters. These results give strong evidence that separating the network into parallel uncertainty and prediction streams leads to state-of-the-art performance with accurate uncertainty estimates.