Learning 3D Semantic Segmentation with only 2D Image Supervision

by   Kyle Genova, et al.

With the recent growth of urban mapping and autonomous driving efforts, there has been an explosion of raw 3D data collected from terrestrial platforms with lidar scanners and color cameras. However, due to high labeling costs, ground-truth 3D semantic segmentation annotations are limited in both quantity and geographic diversity, while also being difficult to transfer across sensors. In contrast, large image collections with ground-truth semantic segmentations are readily available for diverse sets of scenes. In this paper, we investigate how to use only those labeled 2D image collections to supervise training 3D semantic segmentation models. Our approach is to train a 3D model from pseudo-labels derived from 2D semantic image segmentations using multiview fusion. We address several novel issues with this approach, including how to select trusted pseudo-labels, how to sample 3D scenes with rare object categories, and how to decouple input features from 2D images from pseudo-labels during training. The proposed network architecture, 2D3DNet, achieves significantly better performance (+6.2-11.4 mIoU) than baselines during experiments on a new urban dataset with lidar and images captured in 20 cities across 5 continents.



There are no comments yet.


page 1

page 3

page 4

page 7

page 11

page 12


Semantic Segmentation on Swiss3DCities: A Benchmark Study on Aerial Photogrammetric 3D Pointcloud Dataset

We introduce a new outdoor urban 3D pointcloud dataset, covering a total...

PillarSegNet: Pillar-based Semantic Grid Map Estimation using Sparse LiDAR Data

Semantic understanding of the surrounding environment is essential for a...

Can Ground Truth Label Propagation from Video help Semantic Segmentation?

For state-of-the-art semantic segmentation task, training convolutional ...

A Bayesian Approach to Reinforcement Learning of Vision-Based Vehicular Control

In this paper, we present a state-of-the-art reinforcement learning meth...

SemanticPOSS: A Point Cloud Dataset with Large Quantity of Dynamic Instances

3D semantic segmentation is one of the key tasks for autonomous driving ...

Rainy screens: Collecting rainy datasets, indoors

Acquisition of data with adverse conditions in robotics is a cumbersome ...

Multi-domain semantic segmentation with overlapping labels

Deep supervised models have an unprecedented capacity to absorb large qu...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.