GateAttentionPose: Enhancing Pose Estimation with Agent Attention and Improved Gated Convolutions
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913498482606080 |
|---|---|
| author | Feng, Liang Shen, Zhixuan Wen, Lihua Li, Shiyao Xu, Ming |
| author_facet | Feng, Liang Shen, Zhixuan Wen, Lihua Li, Shiyao Xu, Ming |
| contents | This paper introduces GateAttentionPose, an innovative approach that enhances the UniRepLKNet architecture for pose estimation tasks. We present two key contributions: the Agent Attention module and the Gate-Enhanced Feedforward Block (GEFB). The Agent Attention module replaces large kernel convolutions, significantly improving computational efficiency while preserving global context modeling. The GEFB augments feature extraction and processing capabilities, particularly in complex scenes. Extensive evaluations on COCO and MPII datasets demonstrate that GateAttentionPose outperforms existing state-of-the-art methods, including the original UniRepLKNet, achieving superior or comparable results with improved efficiency. Our approach offers a robust solution for pose estimation across diverse applications, including autonomous driving, human motion capture, and virtual reality. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_07798 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | GateAttentionPose: Enhancing Pose Estimation with Agent Attention and Improved Gated Convolutions Feng, Liang Shen, Zhixuan Wen, Lihua Li, Shiyao Xu, Ming Computer Vision and Pattern Recognition This paper introduces GateAttentionPose, an innovative approach that enhances the UniRepLKNet architecture for pose estimation tasks. We present two key contributions: the Agent Attention module and the Gate-Enhanced Feedforward Block (GEFB). The Agent Attention module replaces large kernel convolutions, significantly improving computational efficiency while preserving global context modeling. The GEFB augments feature extraction and processing capabilities, particularly in complex scenes. Extensive evaluations on COCO and MPII datasets demonstrate that GateAttentionPose outperforms existing state-of-the-art methods, including the original UniRepLKNet, achieving superior or comparable results with improved efficiency. Our approach offers a robust solution for pose estimation across diverse applications, including autonomous driving, human motion capture, and virtual reality. |
| title | GateAttentionPose: Enhancing Pose Estimation with Agent Attention and Improved Gated Convolutions |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2409.07798 |