Litcius/Paper detail

Deep Adaptive LiDAR: End-to-end Optimization of Sampling and Depth Completion at Low Sampling Rates

Alexander W. Bergman, David B. Lindell, Gordon Wetzstein

202037 citationsDOI

Abstract

Current LiDAR systems are limited in their ability to capture dense 3D point clouds. To overcome this challenge, deep learning-based depth completion algorithms have been developed to inpaint missing depth guided by an RGB image. However, these methods fail for low sampling rates. Here, we propose an adaptive sampling scheme for LiDAR systems that demonstrates state-of-the-art performance for depth completion at low sampling rates. Our system is fully differentiable, allowing the sparse depth sampling and the depth inpainting components to be trained end-to-end with an upstream task.

Topics & Concepts

End-to-end principleSampling (signal processing)Computer scienceLidarInpaintingArtificial intelligenceComputer visionDepth mapPoint cloudDeep learningImage (mathematics)Remote sensingGeologyFilter (signal processing)Advanced Vision and ImagingAdvanced Optical Sensing TechnologiesOptical measurement and interference techniques
Deep Adaptive LiDAR: End-to-end Optimization of Sampling and Depth Completion at Low Sampling Rates | Litcius