A fused score computation approach to reflect the overlap between the predicted box and the ground truth in pedestrian detection
Abstract In pedestrian detection task, numerous predicted boxes and their corresponding scores are generated and these scores are used to filter these predicted boxes by non‐maximum suppression. This paper analysed the training process of the popular anchor‐based pedestrian detection models (e.g. YO...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2024-11-01
|
| Series: | IET Image Processing |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/ipr2.13251 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846169274938294272 |
|---|---|
| author | Zefei Chen Yongjie Lin Jianmin Xu Kai Lu Zihao Huang |
| author_facet | Zefei Chen Yongjie Lin Jianmin Xu Kai Lu Zihao Huang |
| author_sort | Zefei Chen |
| collection | DOAJ |
| description | Abstract In pedestrian detection task, numerous predicted boxes and their corresponding scores are generated and these scores are used to filter these predicted boxes by non‐maximum suppression. This paper analysed the training process of the popular anchor‐based pedestrian detection models (e.g. YOLO and Faster RCNN), and found that the score of the predicted box reflects the overlap between the corresponding anchor and the ground truth, rather than the predicted box itself. Due to the many‐to‐one strategy adopted by anchor‐based methods, multiple predicted boxes could be generated around one predicted box. This study refers to the number of other predicted boxes around the target predicted box as its local density. When a predicted box has a higher local density, it should have a greater overlap with the ground truth. Therefore, this study proposed the fused score by introducing local density into the score. The experiments showed that replacing the score with the fused score can effectively improve the model's detection accuracy. The code and experiments will soon be open‐sourced at https://github.com/zefeichen/FusedScore. |
| format | Article |
| id | doaj-art-7a5bf0e931654c2ab290fbd557554ac8 |
| institution | Kabale University |
| issn | 1751-9659 1751-9667 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Image Processing |
| spelling | doaj-art-7a5bf0e931654c2ab290fbd557554ac82024-11-13T04:42:31ZengWileyIET Image Processing1751-96591751-96672024-11-0118134287429610.1049/ipr2.13251A fused score computation approach to reflect the overlap between the predicted box and the ground truth in pedestrian detectionZefei Chen0Yongjie Lin1Jianmin Xu2Kai Lu3Zihao Huang4School of Civil Engineering & Transportation South China University of Technology Guangzhou Guangdong Province ChinaSchool of Civil Engineering & Transportation South China University of Technology Guangzhou Guangdong Province ChinaSchool of Civil Engineering & Transportation South China University of Technology Guangzhou Guangdong Province ChinaSchool of Civil Engineering & Transportation South China University of Technology Guangzhou Guangdong Province ChinaSchool of Civil Engineering & Transportation South China University of Technology Guangzhou Guangdong Province ChinaAbstract In pedestrian detection task, numerous predicted boxes and their corresponding scores are generated and these scores are used to filter these predicted boxes by non‐maximum suppression. This paper analysed the training process of the popular anchor‐based pedestrian detection models (e.g. YOLO and Faster RCNN), and found that the score of the predicted box reflects the overlap between the corresponding anchor and the ground truth, rather than the predicted box itself. Due to the many‐to‐one strategy adopted by anchor‐based methods, multiple predicted boxes could be generated around one predicted box. This study refers to the number of other predicted boxes around the target predicted box as its local density. When a predicted box has a higher local density, it should have a greater overlap with the ground truth. Therefore, this study proposed the fused score by introducing local density into the score. The experiments showed that replacing the score with the fused score can effectively improve the model's detection accuracy. The code and experiments will soon be open‐sourced at https://github.com/zefeichen/FusedScore.https://doi.org/10.1049/ipr2.13251aggregationcomputer visiondata analysisdata communicationdata mining |
| spellingShingle | Zefei Chen Yongjie Lin Jianmin Xu Kai Lu Zihao Huang A fused score computation approach to reflect the overlap between the predicted box and the ground truth in pedestrian detection IET Image Processing aggregation computer vision data analysis data communication data mining |
| title | A fused score computation approach to reflect the overlap between the predicted box and the ground truth in pedestrian detection |
| title_full | A fused score computation approach to reflect the overlap between the predicted box and the ground truth in pedestrian detection |
| title_fullStr | A fused score computation approach to reflect the overlap between the predicted box and the ground truth in pedestrian detection |
| title_full_unstemmed | A fused score computation approach to reflect the overlap between the predicted box and the ground truth in pedestrian detection |
| title_short | A fused score computation approach to reflect the overlap between the predicted box and the ground truth in pedestrian detection |
| title_sort | fused score computation approach to reflect the overlap between the predicted box and the ground truth in pedestrian detection |
| topic | aggregation computer vision data analysis data communication data mining |
| url | https://doi.org/10.1049/ipr2.13251 |
| work_keys_str_mv | AT zefeichen afusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection AT yongjielin afusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection AT jianminxu afusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection AT kailu afusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection AT zihaohuang afusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection AT zefeichen fusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection AT yongjielin fusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection AT jianminxu fusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection AT kailu fusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection AT zihaohuang fusedscorecomputationapproachtoreflecttheoverlapbetweenthepredictedboxandthegroundtruthinpedestriandetection |