search

vzyrianov / lidargen

Official implementation of "Learning to Generate Realistic LiDAR Point Clouds" (ECCV 2022)
https://www.zyrianov.org/lidargen/
MIT License
135 stars 14 forks source link

Question about Unsupervised LiDAR Densification #3

Open sky-fly97 opened 2 years ago

sky-fly97 commented 2 years ago

Hello, I have two questions to ask for advice.

  1. In simple terms, is the task of this article to upsample a given point cloud? From one input can be sampled to get different styles of virtaual point clouds, thus expanding the dataset?
  2. How to understand Unsupervised LiDAR Densification, with a 4-beam point cloud as input, and supervision is 16-beam, so where is the unsupervised representation?
vzyrianov commented 1 year ago

Hello Xiangchao, thank you for checking out our paper!

  1. The task of the article is to learn the distribution of LiDAR data with a diffusion model. Once the distribution is learned, it is possible to generate unconditional samples or use guided diffusion for conditional sampling (e.g., conditioning on 4 beams to upsample to 64 beam).

  2. There is no supervision for the densification task. Only the sampling process is modified. In the results, we upsample 4 -> 64 and 16 -> 64 by using a form of classifier guidance.