Domain Adaptive Subterranean 3D Pedestrian Detection via Instance Transfer and Confidence Guidance
With the exploration of subterranean scenes, determining how to ensure the safety of subterranean pedestrians has gradually become a hot research topic. Considering the poor illumination and lack of annotated data in subterranean scenes, it is essential to explore the LiDAR-based domain adaptive det...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2024-03-01
|
Series: | Electronics |
Subjects: | |
Online Access: | https://www.mdpi.com/2079-9292/13/5/982 |
_version_ | 1797264630366601216 |
---|---|
author | Zengyun Liu Zexun Zheng Tianyi Qin Liying Xu Xu Zhang |
author_facet | Zengyun Liu Zexun Zheng Tianyi Qin Liying Xu Xu Zhang |
author_sort | Zengyun Liu |
collection | DOAJ |
description | With the exploration of subterranean scenes, determining how to ensure the safety of subterranean pedestrians has gradually become a hot research topic. Considering the poor illumination and lack of annotated data in subterranean scenes, it is essential to explore the LiDAR-based domain adaptive detectors for localizing the spatial location of pedestrians, thus providing instruction for evacuation and rescue. In this paper, a novel domain adaptive subterranean 3D pedestrian detection method is proposed to adapt pre-trained detectors from the annotated road scenes to the unannotated subterranean scenes. Specifically, an instance transfer-based scene updating strategy is designed to update the subterranean scenes by transferring instances from the road scenes to the subterranean scenes, aiming to create sufficient high-quality pseudo labels for fine-tuning the pre-trained detector. In addition, a pseudo label confidence-guided learning mechanism is constructed to fully utilize pseudo labels of different qualities under the guidance of confidence scores. Extensive experiments validate the superiority of our proposed domain adaptive subterranean 3D pedestrian detection method. |
first_indexed | 2024-04-25T00:31:57Z |
format | Article |
id | doaj.art-2f1e81ba9ad64d0c8a293e044040235e |
institution | Directory Open Access Journal |
issn | 2079-9292 |
language | English |
last_indexed | 2024-04-25T00:31:57Z |
publishDate | 2024-03-01 |
publisher | MDPI AG |
record_format | Article |
series | Electronics |
spelling | doaj.art-2f1e81ba9ad64d0c8a293e044040235e2024-03-12T16:42:49ZengMDPI AGElectronics2079-92922024-03-0113598210.3390/electronics13050982Domain Adaptive Subterranean 3D Pedestrian Detection via Instance Transfer and Confidence GuidanceZengyun Liu0Zexun Zheng1Tianyi Qin2Liying Xu3Xu Zhang4School of Electrical and Information Engineering, Tianjin University, Tianjin 300072, ChinaSchool of Electrical and Information Engineering, Tianjin University, Tianjin 300072, ChinaSchool of Electrical and Information Engineering, Tianjin University, Tianjin 300072, ChinaSchool of Electrical and Information Engineering, Tianjin University, Tianjin 300072, ChinaSchool of Electrical and Information Engineering, Tianjin University, Tianjin 300072, ChinaWith the exploration of subterranean scenes, determining how to ensure the safety of subterranean pedestrians has gradually become a hot research topic. Considering the poor illumination and lack of annotated data in subterranean scenes, it is essential to explore the LiDAR-based domain adaptive detectors for localizing the spatial location of pedestrians, thus providing instruction for evacuation and rescue. In this paper, a novel domain adaptive subterranean 3D pedestrian detection method is proposed to adapt pre-trained detectors from the annotated road scenes to the unannotated subterranean scenes. Specifically, an instance transfer-based scene updating strategy is designed to update the subterranean scenes by transferring instances from the road scenes to the subterranean scenes, aiming to create sufficient high-quality pseudo labels for fine-tuning the pre-trained detector. In addition, a pseudo label confidence-guided learning mechanism is constructed to fully utilize pseudo labels of different qualities under the guidance of confidence scores. Extensive experiments validate the superiority of our proposed domain adaptive subterranean 3D pedestrian detection method.https://www.mdpi.com/2079-9292/13/5/982pedestrian detectionsubterranean scenesdomain adaptationpoint cloud |
spellingShingle | Zengyun Liu Zexun Zheng Tianyi Qin Liying Xu Xu Zhang Domain Adaptive Subterranean 3D Pedestrian Detection via Instance Transfer and Confidence Guidance Electronics pedestrian detection subterranean scenes domain adaptation point cloud |
title | Domain Adaptive Subterranean 3D Pedestrian Detection via Instance Transfer and Confidence Guidance |
title_full | Domain Adaptive Subterranean 3D Pedestrian Detection via Instance Transfer and Confidence Guidance |
title_fullStr | Domain Adaptive Subterranean 3D Pedestrian Detection via Instance Transfer and Confidence Guidance |
title_full_unstemmed | Domain Adaptive Subterranean 3D Pedestrian Detection via Instance Transfer and Confidence Guidance |
title_short | Domain Adaptive Subterranean 3D Pedestrian Detection via Instance Transfer and Confidence Guidance |
title_sort | domain adaptive subterranean 3d pedestrian detection via instance transfer and confidence guidance |
topic | pedestrian detection subterranean scenes domain adaptation point cloud |
url | https://www.mdpi.com/2079-9292/13/5/982 |
work_keys_str_mv | AT zengyunliu domainadaptivesubterranean3dpedestriandetectionviainstancetransferandconfidenceguidance AT zexunzheng domainadaptivesubterranean3dpedestriandetectionviainstancetransferandconfidenceguidance AT tianyiqin domainadaptivesubterranean3dpedestriandetectionviainstancetransferandconfidenceguidance AT liyingxu domainadaptivesubterranean3dpedestriandetectionviainstancetransferandconfidenceguidance AT xuzhang domainadaptivesubterranean3dpedestriandetectionviainstancetransferandconfidenceguidance |