Boundary distribution estimation for precise object detection
In the field of state-of-the-art object detection, the task of object localization is typically accomplished through a dedicated subnet that emphasizes bounding box regression. This subnet traditionally predicts the object's position by regressing the box's center position and scaling fact...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIMS Press
2023-07-01
|
| Series: | Electronic Research Archive |
| Subjects: | |
| Online Access: | https://www.aimspress.com/article/doi/10.3934/era.2023257?viewType=HTML |
| _version_ | 1797690930543722496 |
|---|---|
| author | Peng Zhi Haoran Zhou Hang Huang Rui Zhao Rui Zhou Qingguo Zhou |
| author_facet | Peng Zhi Haoran Zhou Hang Huang Rui Zhao Rui Zhou Qingguo Zhou |
| author_sort | Peng Zhi |
| collection | DOAJ |
| description | In the field of state-of-the-art object detection, the task of object localization is typically accomplished through a dedicated subnet that emphasizes bounding box regression. This subnet traditionally predicts the object's position by regressing the box's center position and scaling factors. Despite the widespread adoption of this approach, we have observed that the localization results often suffer from defects, leading to unsatisfactory detector performance. In this paper, we address the shortcomings of previous methods through theoretical analysis and experimental verification and present an innovative solution for precise object detection. Instead of solely focusing on the object's center and size, our approach enhances the accuracy of bounding box localization by refining the box edges based on the estimated distribution at the object's boundary. Experimental results demonstrate the potential and generalizability of our proposed method. |
| first_indexed | 2024-03-12T02:07:13Z |
| format | Article |
| id | doaj.art-74f1b83d7d7245bcbe16ac3dc926f392 |
| institution | Directory Open Access Journal |
| issn | 2688-1594 |
| language | English |
| last_indexed | 2024-03-12T02:07:13Z |
| publishDate | 2023-07-01 |
| publisher | AIMS Press |
| record_format | Article |
| series | Electronic Research Archive |
| spelling | doaj.art-74f1b83d7d7245bcbe16ac3dc926f3922023-09-07T03:31:25ZengAIMS PressElectronic Research Archive2688-15942023-07-013185025503810.3934/era.2023257Boundary distribution estimation for precise object detectionPeng Zhi 0Haoran Zhou 1Hang Huang2Rui Zhao3Rui Zhou 4Qingguo Zhou5School of Information Science and Engineering, Lanzhou University, Lanzhou, ChinaSchool of Information Science and Engineering, Lanzhou University, Lanzhou, ChinaSchool of Information Science and Engineering, Lanzhou University, Lanzhou, ChinaSchool of Information Science and Engineering, Lanzhou University, Lanzhou, ChinaSchool of Information Science and Engineering, Lanzhou University, Lanzhou, ChinaSchool of Information Science and Engineering, Lanzhou University, Lanzhou, ChinaIn the field of state-of-the-art object detection, the task of object localization is typically accomplished through a dedicated subnet that emphasizes bounding box regression. This subnet traditionally predicts the object's position by regressing the box's center position and scaling factors. Despite the widespread adoption of this approach, we have observed that the localization results often suffer from defects, leading to unsatisfactory detector performance. In this paper, we address the shortcomings of previous methods through theoretical analysis and experimental verification and present an innovative solution for precise object detection. Instead of solely focusing on the object's center and size, our approach enhances the accuracy of bounding box localization by refining the box edges based on the estimated distribution at the object's boundary. Experimental results demonstrate the potential and generalizability of our proposed method.https://www.aimspress.com/article/doi/10.3934/era.2023257?viewType=HTMLobject detectiondeep learningboundary estimationbox refinement |
| spellingShingle | Peng Zhi Haoran Zhou Hang Huang Rui Zhao Rui Zhou Qingguo Zhou Boundary distribution estimation for precise object detection Electronic Research Archive object detection deep learning boundary estimation box refinement |
| title | Boundary distribution estimation for precise object detection |
| title_full | Boundary distribution estimation for precise object detection |
| title_fullStr | Boundary distribution estimation for precise object detection |
| title_full_unstemmed | Boundary distribution estimation for precise object detection |
| title_short | Boundary distribution estimation for precise object detection |
| title_sort | boundary distribution estimation for precise object detection |
| topic | object detection deep learning boundary estimation box refinement |
| url | https://www.aimspress.com/article/doi/10.3934/era.2023257?viewType=HTML |
| work_keys_str_mv | AT pengzhi boundarydistributionestimationforpreciseobjectdetection AT haoranzhou boundarydistributionestimationforpreciseobjectdetection AT hanghuang boundarydistributionestimationforpreciseobjectdetection AT ruizhao boundarydistributionestimationforpreciseobjectdetection AT ruizhou boundarydistributionestimationforpreciseobjectdetection AT qingguozhou boundarydistributionestimationforpreciseobjectdetection |