Depth-Independent Depth Completion via Least Square Estimation

TL;DR

This paper introduces a novel depth completion method that uses least squares estimation to independently process RGB images and sparse depth maps, enabling robust, high-quality, super-resolution depth maps with good generalization across datasets.

Contribution

The proposed approach decouples neural network processing from sparse depth input using least squares, allowing flexible handling of varying sparsity and enabling super-resolution depth map generation.

Findings

Achieves competitive performance on benchmark datasets.

Effectively handles varying sparsity in depth data.

Produces high-quality super-resolution depth maps.

Abstract

The depth completion task aims to complete a per-pixel dense depth map from a sparse depth map. In this paper, we propose an efficient least square based depth-independent method to complete the sparse depth map utilizing the RGB image and the sparse depth map in two independent stages. In this way can we decouple the neural network and the sparse depth input, so that when some features of the sparse depth map change, such as the sparsity, our method can still produce a promising result. Moreover, due to the positional encoding and linear procession in our pipeline, we can easily produce a super-resolution dense depth map of high quality. We also test the generalization of our method on different datasets compared to some state-of-the-art algorithms. Experiments on the benchmark show that our method produces competitive performance.