From CAD Models to Soft Point Cloud Labels: An Automatic Annotation Pipeline for Cheaply Supervised 3D Semantic Segmentation
We propose a fully automatic annotation scheme that takes a raw 3D point cloud with a set of fitted CAD models as input and outputs convincing point-wise labels that can be used as cheap training data for point cloud segmentation. Compared with manual annotations, we show that our automatic labels a...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-07-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/14/3578 |
| _version_ | 1827731866910720000 |
|---|---|
| author | Galadrielle Humblot-Renaux Simon Buus Jensen Andreas Møgelmose |
| author_facet | Galadrielle Humblot-Renaux Simon Buus Jensen Andreas Møgelmose |
| author_sort | Galadrielle Humblot-Renaux |
| collection | DOAJ |
| description | We propose a fully automatic annotation scheme that takes a raw 3D point cloud with a set of fitted CAD models as input and outputs convincing point-wise labels that can be used as cheap training data for point cloud segmentation. Compared with manual annotations, we show that our automatic labels are accurate while drastically reducing the annotation time and eliminating the need for manual intervention or dataset-specific parameters. Our labeling pipeline outputs semantic classes and soft point-wise object scores, which can either be binarized into standard one-hot-encoded labels, thresholded into weak labels with ambiguous points left unlabeled, or used directly as soft labels during training. We evaluate the label quality and segmentation performance of PointNet++ on a dataset of real industrial point clouds and Scan2CAD, a public dataset of indoor scenes. Our results indicate that reducing supervision in areas that are more difficult to label automatically is beneficial compared with the conventional approach of naively assigning a hard “best guess” label to every point. |
| first_indexed | 2024-03-11T00:40:45Z |
| format | Article |
| id | doaj.art-2459c4f649184e2a8373da2395c622a8 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-11T00:40:45Z |
| publishDate | 2023-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-2459c4f649184e2a8373da2395c622a82023-11-18T21:12:45ZengMDPI AGRemote Sensing2072-42922023-07-011514357810.3390/rs15143578From CAD Models to Soft Point Cloud Labels: An Automatic Annotation Pipeline for Cheaply Supervised 3D Semantic SegmentationGaladrielle Humblot-Renaux0Simon Buus Jensen1Andreas Møgelmose2Visual Analysis and Perception Lab, Aalborg University, 9000 Aalborg, DenmarkVisual Analysis and Perception Lab, Aalborg University, 9000 Aalborg, DenmarkVisual Analysis and Perception Lab, Aalborg University, 9000 Aalborg, DenmarkWe propose a fully automatic annotation scheme that takes a raw 3D point cloud with a set of fitted CAD models as input and outputs convincing point-wise labels that can be used as cheap training data for point cloud segmentation. Compared with manual annotations, we show that our automatic labels are accurate while drastically reducing the annotation time and eliminating the need for manual intervention or dataset-specific parameters. Our labeling pipeline outputs semantic classes and soft point-wise object scores, which can either be binarized into standard one-hot-encoded labels, thresholded into weak labels with ambiguous points left unlabeled, or used directly as soft labels during training. We evaluate the label quality and segmentation performance of PointNet++ on a dataset of real industrial point clouds and Scan2CAD, a public dataset of indoor scenes. Our results indicate that reducing supervision in areas that are more difficult to label automatically is beneficial compared with the conventional approach of naively assigning a hard “best guess” label to every point.https://www.mdpi.com/2072-4292/15/14/35783D semantic segmentationautomatic labelingsoft labelspoint cloudsdeep learning |
| spellingShingle | Galadrielle Humblot-Renaux Simon Buus Jensen Andreas Møgelmose From CAD Models to Soft Point Cloud Labels: An Automatic Annotation Pipeline for Cheaply Supervised 3D Semantic Segmentation Remote Sensing 3D semantic segmentation automatic labeling soft labels point clouds deep learning |
| title | From CAD Models to Soft Point Cloud Labels: An Automatic Annotation Pipeline for Cheaply Supervised 3D Semantic Segmentation |
| title_full | From CAD Models to Soft Point Cloud Labels: An Automatic Annotation Pipeline for Cheaply Supervised 3D Semantic Segmentation |
| title_fullStr | From CAD Models to Soft Point Cloud Labels: An Automatic Annotation Pipeline for Cheaply Supervised 3D Semantic Segmentation |
| title_full_unstemmed | From CAD Models to Soft Point Cloud Labels: An Automatic Annotation Pipeline for Cheaply Supervised 3D Semantic Segmentation |
| title_short | From CAD Models to Soft Point Cloud Labels: An Automatic Annotation Pipeline for Cheaply Supervised 3D Semantic Segmentation |
| title_sort | from cad models to soft point cloud labels an automatic annotation pipeline for cheaply supervised 3d semantic segmentation |
| topic | 3D semantic segmentation automatic labeling soft labels point clouds deep learning |
| url | https://www.mdpi.com/2072-4292/15/14/3578 |
| work_keys_str_mv | AT galadriellehumblotrenaux fromcadmodelstosoftpointcloudlabelsanautomaticannotationpipelineforcheaplysupervised3dsemanticsegmentation AT simonbuusjensen fromcadmodelstosoftpointcloudlabelsanautomaticannotationpipelineforcheaplysupervised3dsemanticsegmentation AT andreasmøgelmose fromcadmodelstosoftpointcloudlabelsanautomaticannotationpipelineforcheaplysupervised3dsemanticsegmentation |