# Deep Generative Modeling of LiDAR Data

## TL;DR

This paper introduces a deep generative model for LiDAR data that produces high-quality, structured 2D point maps, improving over existing methods and robustly handling noisy inputs by leveraging a novel data representation.

## Contribution

It adapts deep generative models to LiDAR scan synthesis by unravelling scans into 2D maps and proposes a new data representation that enhances robustness and quality.

## Key findings

- Significant improvements over state-of-the-art point cloud generation methods.
- The model can recover LiDAR scans from noisy or incomplete data.
- The proposed data representation improves robustness to input noise.

## Abstract

Building models capable of generating structured output is a key challenge for AI and robotics. While generative models have been explored on many types of data, little work has been done on synthesizing lidar scans, which play a key role in robot mapping and localization. In this work, we show that one can adapt deep generative models for this task by unravelling lidar scans into a 2D point map. Our approach can generate high quality samples, while simultaneously learning a meaningful latent representation of the data. We demonstrate significant improvements against state-of-the-art point cloud generation methods. Furthermore, we propose a novel data representation that augments the 2D signal with absolute positional information. We show that this helps robustness to noisy and imputed input; the learned model can recover the underlying lidar scan from seemingly uninformative data