TKwinFormer: Top k Window Attention in Vision Transformers for Feature Matching
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2023
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909558578872320 |
|---|---|
| author | Liao, Yun Di, Yide Zhou, Hao Zhu, Kaijun Lu, Mingyu Zhang, Yijia Duan, Qing Liu, Junhui |
| author_facet | Liao, Yun Di, Yide Zhou, Hao Zhu, Kaijun Lu, Mingyu Zhang, Yijia Duan, Qing Liu, Junhui |
| contents | Local feature matching remains a challenging task, primarily due to difficulties in matching sparse keypoints and low-texture regions. The key to solving this problem lies in effectively and accurately integrating global and local information. To achieve this goal, we introduce an innovative local feature matching method called TKwinFormer. Our approach employs a multi-stage matching strategy to optimize the efficiency of information interaction. Furthermore, we propose a novel attention mechanism called Top K Window Attention, which facilitates global information interaction through window tokens prior to patch-level matching, resulting in improved matching accuracy. Additionally, we design an attention block to enhance attention between channels. Experimental results demonstrate that TKwinFormer outperforms state-of-the-art methods on various benchmarks. Code is available at: https://github.com/LiaoYun0x0/TKwinFormer. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2308_15144 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | TKwinFormer: Top k Window Attention in Vision Transformers for Feature Matching Liao, Yun Di, Yide Zhou, Hao Zhu, Kaijun Lu, Mingyu Zhang, Yijia Duan, Qing Liu, Junhui Image and Video Processing I.4.7 Local feature matching remains a challenging task, primarily due to difficulties in matching sparse keypoints and low-texture regions. The key to solving this problem lies in effectively and accurately integrating global and local information. To achieve this goal, we introduce an innovative local feature matching method called TKwinFormer. Our approach employs a multi-stage matching strategy to optimize the efficiency of information interaction. Furthermore, we propose a novel attention mechanism called Top K Window Attention, which facilitates global information interaction through window tokens prior to patch-level matching, resulting in improved matching accuracy. Additionally, we design an attention block to enhance attention between channels. Experimental results demonstrate that TKwinFormer outperforms state-of-the-art methods on various benchmarks. Code is available at: https://github.com/LiaoYun0x0/TKwinFormer. |
| title | TKwinFormer: Top k Window Attention in Vision Transformers for Feature Matching |
| topic | Image and Video Processing I.4.7 |
| url | https://arxiv.org/abs/2308.15144 |