Dayong Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Dayong Li has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Linrunjia Liu
College of Cyber Security, Tarim University, Aral 843300, China
1 shared publication - Qiguang Miao
Xidian University
1 shared publication - Shuai Wu
School of Computer Science and Technology, Xidian University, Xi’an 710071, China
1 shared publication
Latest Publications
- Open Access
- 10 April 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (8) , 3738
Towards Accurate Face Detection Under Occlusion, Class Imbalance and Small-Scale Challenges
To address face occlusion, low detection rates of small-scale faces, and sample imbalance in dense visual scenarios, we propose a YOLOv7-based detector with four key improvements: (1) an optimized MPConv module to enhance feature extraction; (2) a novel CFPM to boost sensitivity to occluded samples; (3) an integration of the DyHead block in IDetect to mitigate feature loss from sample imbalance; (4) an SW-SCE loss function with a dual-input network to better detect small faces. Experiments on the WiderFace dataset show that our method improves detection performance by 1.2%, 1.8%, and 3% on the easy, medium, and hard subsets over the baseline. These gains strengthen face detection in dense, challenging environments with heavy occlusion and small-scale targets.
- 10 April 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (8) , 3738
Open AccessTowards Accurate Face Detection Under Occlusion, Class Imbalance and Small-Scale Challenges
To address face occlusion, low detection rates of small-scale faces, and sample imbalance in dense visual scenarios, we propose a YOLOv7-based detector with four key improvements: (1) an optimized MPConv module to enhance feature extraction; (2) a novel CFPM to boost sensitivity to occluded samples; (3) an integration of the DyHead block in IDetect to mitigate feature loss from sample imbalance; (4) an SW-SCE loss function with a dual-input network to better detect small faces. Experiments on the WiderFace dataset show that our method improves detection performance by 1.2%, 1.8%, and 3% on the easy, medium, and hard subsets over the baseline. These gains strengthen face detection in dense, challenging environments with heavy occlusion and small-scale targets.