Archangel: A Hybrid UAV-Based Human Detection Benchmark With Position and Pose Metadata
Learning to detect objects, such as humans, in imagery captured by an unmanned aerial vehicle (UAV) usually suffers from tremendous variations caused by the UAV’s position towards the objects. In addition, existing UAV-based benchmark datasets do not provide adequate dataset metadata, whi...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10196325/ |
| _version_ | 1797676668615131136 |
|---|---|
| author | Yi-Ting Shen Yaesop Lee Heesung Kwon Damon M. Conover Shuvra S. Bhattacharyya Nikolas Vale Joshua D. Gray G. Jeremy Leong Kenneth Evensen Frank Skirlo |
| author_facet | Yi-Ting Shen Yaesop Lee Heesung Kwon Damon M. Conover Shuvra S. Bhattacharyya Nikolas Vale Joshua D. Gray G. Jeremy Leong Kenneth Evensen Frank Skirlo |
| author_sort | Yi-Ting Shen |
| collection | DOAJ |
| description | Learning to detect objects, such as humans, in imagery captured by an unmanned aerial vehicle (UAV) usually suffers from tremendous variations caused by the UAV’s position towards the objects. In addition, existing UAV-based benchmark datasets do not provide adequate dataset metadata, which is essential for precise model diagnosis and learning features invariant to those variations. In this paper, we introduce Archangel, the first UAV-based object detection dataset composed of real and synthetic subsets captured with similar imagining conditions and UAV position and object pose metadata. A series of experiments are carefully designed with a state-of-the-art object detector to demonstrate the benefits of leveraging the metadata during model evaluation. Moreover, several crucial insights involving both real and synthetic data during model optimization are presented. In the end, we discuss the advantages, limitations, and future directions regarding Archangel to highlight its distinct value for the broader machine learning community. |
| first_indexed | 2024-03-11T22:34:18Z |
| format | Article |
| id | doaj.art-26f6b501c3a74f4db92f18ab92d00b23 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-11T22:34:18Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-26f6b501c3a74f4db92f18ab92d00b232023-09-22T23:00:30ZengIEEEIEEE Access2169-35362023-01-0111809588097210.1109/ACCESS.2023.329923510196325Archangel: A Hybrid UAV-Based Human Detection Benchmark With Position and Pose MetadataYi-Ting Shen0https://orcid.org/0000-0002-1167-5535Yaesop Lee1https://orcid.org/0000-0002-9293-4155Heesung Kwon2Damon M. Conover3Shuvra S. Bhattacharyya4https://orcid.org/0000-0001-7719-1106Nikolas Vale5Joshua D. Gray6https://orcid.org/0009-0006-2615-0287G. Jeremy Leong7Kenneth Evensen8Frank Skirlo9ECE Department, University of Maryland, College Park, MD, USAECE Department, University of Maryland, College Park, MD, USADEVCOM Army Research Laboratory (ARL), Adelphi, MD, USADEVCOM Army Research Laboratory (ARL), Adelphi, MD, USAECE Department, University of Maryland, College Park, MD, USADEVCOM Army Research Laboratory (ARL), Adelphi, MD, USAFibertek Inc., Herndon, VA, USADepartment of Energy, Washington, DC, USADefense Threat Reduction Agency (DTRA), Fort Belvoir, VA, USADefense Threat Reduction Agency (DTRA), Fort Belvoir, VA, USALearning to detect objects, such as humans, in imagery captured by an unmanned aerial vehicle (UAV) usually suffers from tremendous variations caused by the UAV’s position towards the objects. In addition, existing UAV-based benchmark datasets do not provide adequate dataset metadata, which is essential for precise model diagnosis and learning features invariant to those variations. In this paper, we introduce Archangel, the first UAV-based object detection dataset composed of real and synthetic subsets captured with similar imagining conditions and UAV position and object pose metadata. A series of experiments are carefully designed with a state-of-the-art object detector to demonstrate the benefits of leveraging the metadata during model evaluation. Moreover, several crucial insights involving both real and synthetic data during model optimization are presented. In the end, we discuss the advantages, limitations, and future directions regarding Archangel to highlight its distinct value for the broader machine learning community.https://ieeexplore.ieee.org/document/10196325/UAV-based object detectionhuman detectionUAV-based benchmark datasetposition metadatasynthetic datamodel optimization |
| spellingShingle | Yi-Ting Shen Yaesop Lee Heesung Kwon Damon M. Conover Shuvra S. Bhattacharyya Nikolas Vale Joshua D. Gray G. Jeremy Leong Kenneth Evensen Frank Skirlo Archangel: A Hybrid UAV-Based Human Detection Benchmark With Position and Pose Metadata IEEE Access UAV-based object detection human detection UAV-based benchmark dataset position metadata synthetic data model optimization |
| title | Archangel: A Hybrid UAV-Based Human Detection Benchmark With Position and Pose Metadata |
| title_full | Archangel: A Hybrid UAV-Based Human Detection Benchmark With Position and Pose Metadata |
| title_fullStr | Archangel: A Hybrid UAV-Based Human Detection Benchmark With Position and Pose Metadata |
| title_full_unstemmed | Archangel: A Hybrid UAV-Based Human Detection Benchmark With Position and Pose Metadata |
| title_short | Archangel: A Hybrid UAV-Based Human Detection Benchmark With Position and Pose Metadata |
| title_sort | archangel a hybrid uav based human detection benchmark with position and pose metadata |
| topic | UAV-based object detection human detection UAV-based benchmark dataset position metadata synthetic data model optimization |
| url | https://ieeexplore.ieee.org/document/10196325/ |
| work_keys_str_mv | AT yitingshen archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata AT yaesoplee archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata AT heesungkwon archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata AT damonmconover archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata AT shuvrasbhattacharyya archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata AT nikolasvale archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata AT joshuadgray archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata AT gjeremyleong archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata AT kennethevensen archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata AT frankskirlo archangelahybriduavbasedhumandetectionbenchmarkwithpositionandposemetadata |