Pedestrian Detection by Novel Axis-Line Representation and Regression Pattern
The pattern of bounding box representation and regression has long been dominant in CNN-based pedestrian detectors. Despite the method’s success, it cannot accurately represent location, and introduces unnecessary background information, while pedestrian features are mainly located in axis-line area...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-05-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/10/3312 |
| _version_ | 1797534685427924992 |
|---|---|
| author | Mengxue Zhang Qiong Liu |
| author_facet | Mengxue Zhang Qiong Liu |
| author_sort | Mengxue Zhang |
| collection | DOAJ |
| description | The pattern of bounding box representation and regression has long been dominant in CNN-based pedestrian detectors. Despite the method’s success, it cannot accurately represent location, and introduces unnecessary background information, while pedestrian features are mainly located in axis-line areas. Other object representations, such as corner-pairs, are not easy to obtain by regression because the corners are far from the axis-line and are greatly affected by background features. In this paper, we propose a novel detection pattern, named Axis-line Representation and Regression (ALR), for pedestrian detection in road scenes. Specifically, we design a 3-d axis-line representation for pedestrians and use it as the regression target during network training. A line-box transformation method is also proposed to fit the widely used box-annotations. Meanwhile, we explore the influence of deformable convolution base-offset on detection performance and propose a base-offset initialization strategy to further promote the gain brought by ALR. Notably, the proposed ALR pattern can be introduced into both anchor-based and anchor-free frameworks. We validate the effectiveness of ALR on the Caltech-USA and CityPersons datasets. Experimental results show that our approach outperforms the baseline significantly through simple modifications and achieves competitive accuracy with other methods without bells and whistles. |
| first_indexed | 2024-03-10T11:32:58Z |
| format | Article |
| id | doaj.art-aa541a1bfa7b4b958b2d3d0226aa7d12 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T11:32:58Z |
| publishDate | 2021-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-aa541a1bfa7b4b958b2d3d0226aa7d122023-11-21T19:06:42ZengMDPI AGSensors1424-82202021-05-012110331210.3390/s21103312Pedestrian Detection by Novel Axis-Line Representation and Regression PatternMengxue Zhang0Qiong Liu1School of Software Engineering, South China University of Technology, Guangzhou 510006, ChinaSchool of Software Engineering, South China University of Technology, Guangzhou 510006, ChinaThe pattern of bounding box representation and regression has long been dominant in CNN-based pedestrian detectors. Despite the method’s success, it cannot accurately represent location, and introduces unnecessary background information, while pedestrian features are mainly located in axis-line areas. Other object representations, such as corner-pairs, are not easy to obtain by regression because the corners are far from the axis-line and are greatly affected by background features. In this paper, we propose a novel detection pattern, named Axis-line Representation and Regression (ALR), for pedestrian detection in road scenes. Specifically, we design a 3-d axis-line representation for pedestrians and use it as the regression target during network training. A line-box transformation method is also proposed to fit the widely used box-annotations. Meanwhile, we explore the influence of deformable convolution base-offset on detection performance and propose a base-offset initialization strategy to further promote the gain brought by ALR. Notably, the proposed ALR pattern can be introduced into both anchor-based and anchor-free frameworks. We validate the effectiveness of ALR on the Caltech-USA and CityPersons datasets. Experimental results show that our approach outperforms the baseline significantly through simple modifications and achieves competitive accuracy with other methods without bells and whistles.https://www.mdpi.com/1424-8220/21/10/3312pedestrian detectionobject representationaxis lineroad scene |
| spellingShingle | Mengxue Zhang Qiong Liu Pedestrian Detection by Novel Axis-Line Representation and Regression Pattern Sensors pedestrian detection object representation axis line road scene |
| title | Pedestrian Detection by Novel Axis-Line Representation and Regression Pattern |
| title_full | Pedestrian Detection by Novel Axis-Line Representation and Regression Pattern |
| title_fullStr | Pedestrian Detection by Novel Axis-Line Representation and Regression Pattern |
| title_full_unstemmed | Pedestrian Detection by Novel Axis-Line Representation and Regression Pattern |
| title_short | Pedestrian Detection by Novel Axis-Line Representation and Regression Pattern |
| title_sort | pedestrian detection by novel axis line representation and regression pattern |
| topic | pedestrian detection object representation axis line road scene |
| url | https://www.mdpi.com/1424-8220/21/10/3312 |
| work_keys_str_mv | AT mengxuezhang pedestriandetectionbynovelaxislinerepresentationandregressionpattern AT qiongliu pedestriandetectionbynovelaxislinerepresentationandregressionpattern |