Learning 3D Semantic Segmentation with only 2D Image Supervision
This addresses the high labeling costs and limited availability of 3D annotations for applications like urban mapping and autonomous driving, though it is an incremental improvement over existing pseudo-labeling methods.
The paper tackles the problem of training 3D semantic segmentation models without expensive 3D annotations by using only labeled 2D images, achieving a performance gain of +6.2-11.4 mIoU over baselines on a diverse urban dataset.
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.