in , ,

(video) RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds, Qingyong Hu

2nd Workshop 3D-Deep Learning for Autonomous Driving, IV 2020 Las

Title : RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds

Speaker: Qingyong Hu

Abstract : We study the problem of efficient semantic segmentation for large-scale 3D point clouds. By relying on expensive sampling techniques or computationally heavy pre/post-processing steps, most existing approaches are only able to be trained and operate over small-scale point clouds. In this paper, we introduce RandLA-Net, an efficient and lightweight neural architecture to directly infer per-point semantics for large-scale point clouds.

The key to our approach is to use random point sampling instead of more complex point selection approaches. Although remarkably computation and memory efficient, random sampling can discard key features by chance. To overcome this, we introduce a novel local feature aggregation module to progressively increase the receptive field for each 3D point, thereby effectively preserving geometric details.

Extensive experiments show that our RandLA-Net can process 1 million points in a single pass with up to 200$times$ faster than existing approaches. Moreover, our RandLA-Net clearly surpasses state-of-the-art approaches for semantic segmentation on two large-scale benchmarks Semantic3D and SemanticKITTI.

Report

What do you think?

487 Points
Upvote Downvote

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

GIPHY App Key not set. Please check settings

(video) Sim2Real: Learning to Drive from Simulation without Real World Labels

Zenzic announces five winners of £500,000 funding to help UK self-driving start-ups

Zenzic announces five winners of £500,000 funding to help UK self-driving start-ups