Exploring 3D Object Detection for Autonomous Factory Driving: Advanced Research on Handling Limited Annotations with Ground Truth Sampling Augmentation
Autonomously driving vehicles in car factories and parking spaces can represent a competitive advantage in the logistics industry. However, the real-world application is challenging in many ways. First of all, there are no publicly available datasets for this specific task. Therefore, we equipped tw...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-02-01
|
| Series: | Computer Sciences & Mathematics Forum |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2813-0324/9/1/5 |
| _version_ | 1797241526216032256 |
|---|---|
| author | Matthias Reuse Karl Amende Martin Simon Bernhard Sick |
| author_facet | Matthias Reuse Karl Amende Martin Simon Bernhard Sick |
| author_sort | Matthias Reuse |
| collection | DOAJ |
| description | Autonomously driving vehicles in car factories and parking spaces can represent a competitive advantage in the logistics industry. However, the real-world application is challenging in many ways. First of all, there are no publicly available datasets for this specific task. Therefore, we equipped two industrial production sites with up to 11 LiDAR sensors to collect and annotate our own data for infrastructural 3D object detection. These form the basis for extensive experiments. Due to the still limited amount of labeled data, the commonly used ground truth sampling augmentation is the core of research in this work. Several variations of this augmentation method are explored, revealing that in our case, the most commonly used is not necessarily the best. We show that an easy-to-create polygon can noticeably improve the detection results in this application scenario. By using these augmentation methods, it is even possible to achieve moderate detection results when only empty frames without any objects and a database with only a few labeled objects are used. |
| first_indexed | 2024-04-24T18:24:43Z |
| format | Article |
| id | doaj.art-8fc5059a16c942ceb411b53a515b2257 |
| institution | Directory Open Access Journal |
| issn | 2813-0324 |
| language | English |
| last_indexed | 2024-04-24T18:24:43Z |
| publishDate | 2024-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Computer Sciences & Mathematics Forum |
| spelling | doaj.art-8fc5059a16c942ceb411b53a515b22572024-03-27T13:32:36ZengMDPI AGComputer Sciences & Mathematics Forum2813-03242024-02-0191510.3390/cmsf2024009005Exploring 3D Object Detection for Autonomous Factory Driving: Advanced Research on Handling Limited Annotations with Ground Truth Sampling AugmentationMatthias Reuse0Karl Amende1Martin Simon2Bernhard Sick3Valeo Schalter & Sensoren GmbH, Hummendorfer Str. 74, 96317 Kronach, GermanyValeo Schalter & Sensoren GmbH, Hummendorfer Str. 74, 96317 Kronach, GermanyValeo Schalter & Sensoren GmbH, Hummendorfer Str. 74, 96317 Kronach, GermanyIntelligent Embedded Systems, Universität Kassel, Wilhelmshöher Allee 73, 34121 Kassel, GermanyAutonomously driving vehicles in car factories and parking spaces can represent a competitive advantage in the logistics industry. However, the real-world application is challenging in many ways. First of all, there are no publicly available datasets for this specific task. Therefore, we equipped two industrial production sites with up to 11 LiDAR sensors to collect and annotate our own data for infrastructural 3D object detection. These form the basis for extensive experiments. Due to the still limited amount of labeled data, the commonly used ground truth sampling augmentation is the core of research in this work. Several variations of this augmentation method are explored, revealing that in our case, the most commonly used is not necessarily the best. We show that an easy-to-create polygon can noticeably improve the detection results in this application scenario. By using these augmentation methods, it is even possible to achieve moderate detection results when only empty frames without any objects and a database with only a few labeled objects are used.https://www.mdpi.com/2813-0324/9/1/53D object detectioninfrastructural LiDARdata augmentationautonomous driving |
| spellingShingle | Matthias Reuse Karl Amende Martin Simon Bernhard Sick Exploring 3D Object Detection for Autonomous Factory Driving: Advanced Research on Handling Limited Annotations with Ground Truth Sampling Augmentation Computer Sciences & Mathematics Forum 3D object detection infrastructural LiDAR data augmentation autonomous driving |
| title | Exploring 3D Object Detection for Autonomous Factory Driving: Advanced Research on Handling Limited Annotations with Ground Truth Sampling Augmentation |
| title_full | Exploring 3D Object Detection for Autonomous Factory Driving: Advanced Research on Handling Limited Annotations with Ground Truth Sampling Augmentation |
| title_fullStr | Exploring 3D Object Detection for Autonomous Factory Driving: Advanced Research on Handling Limited Annotations with Ground Truth Sampling Augmentation |
| title_full_unstemmed | Exploring 3D Object Detection for Autonomous Factory Driving: Advanced Research on Handling Limited Annotations with Ground Truth Sampling Augmentation |
| title_short | Exploring 3D Object Detection for Autonomous Factory Driving: Advanced Research on Handling Limited Annotations with Ground Truth Sampling Augmentation |
| title_sort | exploring 3d object detection for autonomous factory driving advanced research on handling limited annotations with ground truth sampling augmentation |
| topic | 3D object detection infrastructural LiDAR data augmentation autonomous driving |
| url | https://www.mdpi.com/2813-0324/9/1/5 |
| work_keys_str_mv | AT matthiasreuse exploring3dobjectdetectionforautonomousfactorydrivingadvancedresearchonhandlinglimitedannotationswithgroundtruthsamplingaugmentation AT karlamende exploring3dobjectdetectionforautonomousfactorydrivingadvancedresearchonhandlinglimitedannotationswithgroundtruthsamplingaugmentation AT martinsimon exploring3dobjectdetectionforautonomousfactorydrivingadvancedresearchonhandlinglimitedannotationswithgroundtruthsamplingaugmentation AT bernhardsick exploring3dobjectdetectionforautonomousfactorydrivingadvancedresearchonhandlinglimitedannotationswithgroundtruthsamplingaugmentation |