A Task-Risk Consistency Object Detection Framework Based on Deep Reinforcement Learning
A discernible gap has materialized between the expectations for object detection tasks in optical remote sensing images and the increasingly sophisticated design methods. The flexibility of deep learning object detection algorithms allows the selection and combination of multiple basic structures an...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-10-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/20/5031 |
| _version_ | 1797572424324087808 |
|---|---|
| author | Jiazheng Wen Huanyu Liu Junbao Li |
| author_facet | Jiazheng Wen Huanyu Liu Junbao Li |
| author_sort | Jiazheng Wen |
| collection | DOAJ |
| description | A discernible gap has materialized between the expectations for object detection tasks in optical remote sensing images and the increasingly sophisticated design methods. The flexibility of deep learning object detection algorithms allows the selection and combination of multiple basic structures and model sizes, but this selection process relies heavily on human experience and lacks reliability when faced with special scenarios or extreme data distribution. To address these inherent challenges, this study proposes an approach that leverages deep reinforcement learning within the framework of vision tasks. This study introduces a Task-Risk Consistent Intelligent Detection Framework (TRC-ODF) for object detection in optical remote sensing images. The proposed framework designs a model optimization strategy based on deep reinforcement learning that systematically integrates the available information from images and vision processes. The core of the reinforcement learning agent is the proposed task-risk consistency reward mechanism, which is the driving force behind the optimal prediction allocation in the decision-making process. To verify the effectiveness of the proposed framework, multiple sets of empirical evaluations are conducted on representative optical remote sensing image datasets: RSOD, NWPU VHR-10, and DIOR. When applying the proposed framework to representative advanced detection models, the mean average precision (mAP@0.5 and mAP@0.5:0.95) is improved by 0.8–5.4 and 0.4–2.7, respectively. The obtained results showcase the considerable promise and potential of the TRC-ODF framework to address the challenges associated with object detection in optical remote sensing images. |
| first_indexed | 2024-03-10T20:56:05Z |
| format | Article |
| id | doaj.art-d5f4d656e056450d8716271f0e0ac024 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T20:56:05Z |
| publishDate | 2023-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-d5f4d656e056450d8716271f0e0ac0242023-11-19T17:59:52ZengMDPI AGRemote Sensing2072-42922023-10-011520503110.3390/rs15205031A Task-Risk Consistency Object Detection Framework Based on Deep Reinforcement LearningJiazheng Wen0Huanyu Liu1Junbao Li2Faculty of Computing, Harbin Institute of Technology, Harbin 150080, ChinaFaculty of Computing, Harbin Institute of Technology, Harbin 150080, ChinaFaculty of Computing, Harbin Institute of Technology, Harbin 150080, ChinaA discernible gap has materialized between the expectations for object detection tasks in optical remote sensing images and the increasingly sophisticated design methods. The flexibility of deep learning object detection algorithms allows the selection and combination of multiple basic structures and model sizes, but this selection process relies heavily on human experience and lacks reliability when faced with special scenarios or extreme data distribution. To address these inherent challenges, this study proposes an approach that leverages deep reinforcement learning within the framework of vision tasks. This study introduces a Task-Risk Consistent Intelligent Detection Framework (TRC-ODF) for object detection in optical remote sensing images. The proposed framework designs a model optimization strategy based on deep reinforcement learning that systematically integrates the available information from images and vision processes. The core of the reinforcement learning agent is the proposed task-risk consistency reward mechanism, which is the driving force behind the optimal prediction allocation in the decision-making process. To verify the effectiveness of the proposed framework, multiple sets of empirical evaluations are conducted on representative optical remote sensing image datasets: RSOD, NWPU VHR-10, and DIOR. When applying the proposed framework to representative advanced detection models, the mean average precision (mAP@0.5 and mAP@0.5:0.95) is improved by 0.8–5.4 and 0.4–2.7, respectively. The obtained results showcase the considerable promise and potential of the TRC-ODF framework to address the challenges associated with object detection in optical remote sensing images.https://www.mdpi.com/2072-4292/15/20/5031object detectionreinforcement learningoptical remote sensing image |
| spellingShingle | Jiazheng Wen Huanyu Liu Junbao Li A Task-Risk Consistency Object Detection Framework Based on Deep Reinforcement Learning Remote Sensing object detection reinforcement learning optical remote sensing image |
| title | A Task-Risk Consistency Object Detection Framework Based on Deep Reinforcement Learning |
| title_full | A Task-Risk Consistency Object Detection Framework Based on Deep Reinforcement Learning |
| title_fullStr | A Task-Risk Consistency Object Detection Framework Based on Deep Reinforcement Learning |
| title_full_unstemmed | A Task-Risk Consistency Object Detection Framework Based on Deep Reinforcement Learning |
| title_short | A Task-Risk Consistency Object Detection Framework Based on Deep Reinforcement Learning |
| title_sort | task risk consistency object detection framework based on deep reinforcement learning |
| topic | object detection reinforcement learning optical remote sensing image |
| url | https://www.mdpi.com/2072-4292/15/20/5031 |
| work_keys_str_mv | AT jiazhengwen ataskriskconsistencyobjectdetectionframeworkbasedondeepreinforcementlearning AT huanyuliu ataskriskconsistencyobjectdetectionframeworkbasedondeepreinforcementlearning AT junbaoli ataskriskconsistencyobjectdetectionframeworkbasedondeepreinforcementlearning AT jiazhengwen taskriskconsistencyobjectdetectionframeworkbasedondeepreinforcementlearning AT huanyuliu taskriskconsistencyobjectdetectionframeworkbasedondeepreinforcementlearning AT junbaoli taskriskconsistencyobjectdetectionframeworkbasedondeepreinforcementlearning |