Prior based Sampling for Adaptive LiDAR

TL;DR

This paper introduces SampleDepth, a CNN-based adaptive LiDAR sampling method that predicts optimal sampling locations using previous depth data to improve downstream depth completion performance.

Contribution

It presents a novel adaptive sampling strategy for LiDAR using a CNN trained to optimize depth completion, unlike traditional fixed sampling methods.

Findings

SampleDepth improves depth completion accuracy across datasets.

It is effective with different depth completion networks.

SampleDepth adapts sampling based on scene content.

Abstract

We propose SampleDepth, a Convolutional Neural Network (CNN), that is suited for an adaptive LiDAR. Typically,LiDAR sampling strategy is pre-defined, constant and independent of the observed scene. Instead of letting a LiDAR sample the scene in this agnostic fashion, SampleDepth determines, adaptively, where it is best to sample the current frame. To do that, SampleDepth uses depth samples from previous time steps to predict a sampling mask for the current frame. Crucially, SampleDepth is trained to optimize the performance of a depth completion downstream task. SampleDepth is evaluated on two different depth completion networks and two LiDAR datasets, KITTI Depth Completion and the newly introduced synthetic dataset, SHIFT. We show that SampleDepth is effective and suitable for different depth completion downstream tasks.