A Multi-Level Adaptive Lightweight Net for Damaged Road Marking Detection Based on Knowledge Distillation
文献类型：期刊论文
| 作者 | Wang, Junwei3,4,5; Zeng, Xiangqiang1,5; Wang, Yong5; Ren, Xiang5; Wang, Dongliang5; Qu, Wenqiu4,5; Liao, Xiaohan5; Pan, Peifen2 |
| 刊名 | REMOTE SENSING
 |
| 出版日期 | 2024-07-01 |
| 卷号 | 16期号:14页码:2593 |
| 关键词 | remote sensing damaged road marking semantic segmentation deep learning knowledge distillation |
| DOI | 10.3390/rs16142593 |
| 产权排序 | 1 |
| 文献子类 | Article |
| 英文摘要 | To tackle the complexity and limited applicability of high-precision segmentation models for damaged road markings, this study proposes a Multi-level Adaptive Lightweight Network (MALNet) based on knowledge distillation. By incorporating multi-scale dilated convolution and adaptive spatial channel attention fusion modules, the MALNet model significantly enhances the precision, integrity, and robustness of its segmentation branch. Furthermore, it employs an intricate knowledge distillation strategy, channeling rich, layered insights from a teacher model to a student model, thus elevating the latter's segmentation ability. Concurrently, it streamlines the student model by markedly reducing its parameter count and computational demands, culminating in a segmentation network that is both high-performing and pragmatic. Rigorous testing on three distinct data sets for damaged road marking detection-CDM_P (Collective Damaged road Marking-Public), CDM_H (Collective Damaged road Marking-Highways), and CDM_C (Collective Damaged road Marking-Cityroad)-underscores the MALNet model's superior segmentation abilities across all damage types, outperforming competing models in accuracy and completeness. Notably, the MALNet model excels in parameter efficiency, computational economy, and throughput. After distillation, the student model's parameters and computational load decrease to only 31.78% and 27.40% of the teacher model's, respectively, while processing speeds increase to 1.9 times, demonstrating a significant improvement in lightweight design. |
| WOS研究方向 | Environmental Sciences & Ecology ; Geology ; Remote Sensing ; Imaging Science & Photographic Technology |
| WOS记录号 | WOS:001277020400001 |
| 出版者 | MDPI |
| 源URL | [http://ir.igsnrr.ac.cn/handle/311030/206944]  |
| 专题 | 资源与环境信息系统国家重点实验室_外文论文 |
| 通讯作者 | Pan, Peifen |
| 作者单位 | 1.Beijing Normal Univ, Fac Geog Sci, State Key Lab Remote Sensing Sci, Beijing 100091, Peoples R China 2.China Acad Railway Sci Grp Co Ltd, Beijing 100081, Peoples R China 3.Beijing Int Data Exchange, Beijing 100027, Peoples R China 4.Univ Chinese Acad Sci, Beijing 100049, Peoples R China 5.Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, State Key Lab Resources & Environm Informat Syst, Beijing 100101, Peoples R China |
| 推荐引用方式 GB/T 7714 | Wang, Junwei,Zeng, Xiangqiang,Wang, Yong,et al. A Multi-Level Adaptive Lightweight Net for Damaged Road Marking Detection Based on Knowledge Distillation[J]. REMOTE SENSING,2024,16(14):2593. |
| APA | Wang, Junwei.,Zeng, Xiangqiang.,Wang, Yong.,Ren, Xiang.,Wang, Dongliang.,...&Pan, Peifen.(2024).A Multi-Level Adaptive Lightweight Net for Damaged Road Marking Detection Based on Knowledge Distillation.REMOTE SENSING,16(14),2593. |
| MLA | Wang, Junwei,et al."A Multi-Level Adaptive Lightweight Net for Damaged Road Marking Detection Based on Knowledge Distillation".REMOTE SENSING 16.14(2024):2593. |
入库方式： OAI收割
来源：地理科学与资源研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。