MGNet: Monocular Geometric Scene Understanding for Autonomous Driving

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Schön, Markus, Buchholz, Michael, Dietmayer, Klaus
Format: Preprint
Published: 2022
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913725819125760
author Schön, Markus
Buchholz, Michael
Dietmayer, Klaus
author_facet Schön, Markus
Buchholz, Michael
Dietmayer, Klaus
contents We introduce MGNet, a multi-task framework for monocular geometric scene understanding. We define monocular geometric scene understanding as the combination of two known tasks: Panoptic segmentation and self-supervised monocular depth estimation. Panoptic segmentation captures the full scene not only semantically, but also on an instance basis. Self-supervised monocular depth estimation uses geometric constraints derived from the camera measurement model in order to measure depth from monocular video sequences only. To the best of our knowledge, we are the first to propose the combination of these two tasks in one single model. Our model is designed with focus on low latency to provide fast inference in real-time on a single consumer-grade GPU. During deployment, our model produces dense 3D point clouds with instance aware semantic labels from single high-resolution camera images. We evaluate our model on two popular autonomous driving benchmarks, i.e., Cityscapes and KITTI, and show competitive performance among other real-time capable methods. Source code is available at https://github.com/markusschoen/MGNet.
format Preprint
id arxiv_https___arxiv_org_abs_2206_13199
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle MGNet: Monocular Geometric Scene Understanding for Autonomous Driving
Schön, Markus
Buchholz, Michael
Dietmayer, Klaus
Computer Vision and Pattern Recognition
We introduce MGNet, a multi-task framework for monocular geometric scene understanding. We define monocular geometric scene understanding as the combination of two known tasks: Panoptic segmentation and self-supervised monocular depth estimation. Panoptic segmentation captures the full scene not only semantically, but also on an instance basis. Self-supervised monocular depth estimation uses geometric constraints derived from the camera measurement model in order to measure depth from monocular video sequences only. To the best of our knowledge, we are the first to propose the combination of these two tasks in one single model. Our model is designed with focus on low latency to provide fast inference in real-time on a single consumer-grade GPU. During deployment, our model produces dense 3D point clouds with instance aware semantic labels from single high-resolution camera images. We evaluate our model on two popular autonomous driving benchmarks, i.e., Cityscapes and KITTI, and show competitive performance among other real-time capable methods. Source code is available at https://github.com/markusschoen/MGNet.
title MGNet: Monocular Geometric Scene Understanding for Autonomous Driving
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2206.13199