Multi‐level features extraction network with gating mechanism for crowd counting
Abstract Crowd counting is still a practical and challenging problem owing to scale variations and information loss. Most existing methods based on the straightforward fusion of different features from a deep neural network seem to eliminate this limitation. However, these features are difficult to...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2021-12-01
|
| Series: | IET Image Processing |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/ipr2.12304 |
| _version_ | 1811344407100653568 |
|---|---|
| author | Xin Zeng Qiang Guo Haoran Duan Yunpeng Wu |
| author_facet | Xin Zeng Qiang Guo Haoran Duan Yunpeng Wu |
| author_sort | Xin Zeng |
| collection | DOAJ |
| description | Abstract Crowd counting is still a practical and challenging problem owing to scale variations and information loss. Most existing methods based on the straightforward fusion of different features from a deep neural network seem to eliminate this limitation. However, these features are difficult to be fused since they often differ significantly in modality and dimensionality. Unlike previous works, a multi‐level features extraction network with gating mechanism for crowd counting is proposed. Specifically, a multi‐channel gated unit to adaptively extract features in different levels of the network is proposed, which can avoid interference from confusing information. To fully aggregate features via multi‐level fusion, multi‐level features extraction scheme is presented. The multi‐level features extraction network learns to fuse features from multiple levels and reduce false predictions. Extensive experiments and evaluations clearly illustrate that the proposed approach achieves state‐of‐the‐art counting performance against other methods on four mainstream crowd counting benchmarks. |
| first_indexed | 2024-04-13T19:47:46Z |
| format | Article |
| id | doaj.art-0606cda287e344b8832e3ee0ca63100c |
| institution | Directory Open Access Journal |
| issn | 1751-9659 1751-9667 |
| language | English |
| last_indexed | 2024-04-13T19:47:46Z |
| publishDate | 2021-12-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Image Processing |
| spelling | doaj.art-0606cda287e344b8832e3ee0ca63100c2022-12-22T02:32:40ZengWileyIET Image Processing1751-96591751-96672021-12-0115143534354210.1049/ipr2.12304Multi‐level features extraction network with gating mechanism for crowd countingXin Zeng0Qiang Guo1Haoran Duan2Yunpeng Wu3ZhengZhou Vocational College of Finance and Taxation Zhengzhou People's Republic of ChinaSchool of Information Engineering Zhengzhou University Zhengzhou People's Republic of ChinaOpen Lab, School of Computing Newcastle University Newcastle Upon Tyne UKSchool of Information Engineering Zhengzhou University Zhengzhou People's Republic of ChinaAbstract Crowd counting is still a practical and challenging problem owing to scale variations and information loss. Most existing methods based on the straightforward fusion of different features from a deep neural network seem to eliminate this limitation. However, these features are difficult to be fused since they often differ significantly in modality and dimensionality. Unlike previous works, a multi‐level features extraction network with gating mechanism for crowd counting is proposed. Specifically, a multi‐channel gated unit to adaptively extract features in different levels of the network is proposed, which can avoid interference from confusing information. To fully aggregate features via multi‐level fusion, multi‐level features extraction scheme is presented. The multi‐level features extraction network learns to fuse features from multiple levels and reduce false predictions. Extensive experiments and evaluations clearly illustrate that the proposed approach achieves state‐of‐the‐art counting performance against other methods on four mainstream crowd counting benchmarks.https://doi.org/10.1049/ipr2.12304Image recognitionSensor fusionComputer vision and image processing techniquesNeural nets |
| spellingShingle | Xin Zeng Qiang Guo Haoran Duan Yunpeng Wu Multi‐level features extraction network with gating mechanism for crowd counting IET Image Processing Image recognition Sensor fusion Computer vision and image processing techniques Neural nets |
| title | Multi‐level features extraction network with gating mechanism for crowd counting |
| title_full | Multi‐level features extraction network with gating mechanism for crowd counting |
| title_fullStr | Multi‐level features extraction network with gating mechanism for crowd counting |
| title_full_unstemmed | Multi‐level features extraction network with gating mechanism for crowd counting |
| title_short | Multi‐level features extraction network with gating mechanism for crowd counting |
| title_sort | multi level features extraction network with gating mechanism for crowd counting |
| topic | Image recognition Sensor fusion Computer vision and image processing techniques Neural nets |
| url | https://doi.org/10.1049/ipr2.12304 |
| work_keys_str_mv | AT xinzeng multilevelfeaturesextractionnetworkwithgatingmechanismforcrowdcounting AT qiangguo multilevelfeaturesextractionnetworkwithgatingmechanismforcrowdcounting AT haoranduan multilevelfeaturesextractionnetworkwithgatingmechanismforcrowdcounting AT yunpengwu multilevelfeaturesextractionnetworkwithgatingmechanismforcrowdcounting |