A Multi-Perspective Detection Head Aided YOLOv8 Algorithm Based on Adaptive Label Assignment for Semiconductor Laser Chip Defect Detection
We proposed an improved YOLOv8 algorithm. The results show that our algorithm has 2.1ms, 0.2% improvement on the detection time and mAP50. Comparing with YOLOv10 and YOLOv9, the prediction time has improved 11.8ms and 12.7ms.
Authors
- Fu Wang (ORCID: https://orcid.org/0000-0002-5596-4208)
- Qi Zhang (ORCID: https://orcid.org/0000-0001-7343-9645)
- Jianwei Zhou (ORCID: https://orcid.org/0000-0002-1288-5905)
- Yutian Li (ORCID: https://orcid.org/0000-0003-1810-3000)
- Qinghua Tian (ORCID: https://orcid.org/0000-0002-1169-9078)
- Jue Wang (ORCID: https://orcid.org/0000-0001-8297-8827)
- Feng Tian
- Biao Luo
Institutions
- Beijing University of Posts and Telecommunications (CN)
Publication Details
- Published
- 2024-11-02
- DOI
- https://doi.org/10.1109/acp/ipoc63121.2024.10809578
- Primary Topic
- Industrial Vision Systems and Defect Detection
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Natural Science Foundation of China
- National Key Research and Development Program of China