Efficient Attention: Attention with Linear Complexities
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2018
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910301363896320 |
|---|---|
| author | Shen, Zhuoran Zhang, Mingyuan Zhao, Haiyu Yi, Shuai Li, Hongsheng |
| author_facet | Shen, Zhuoran Zhang, Mingyuan Zhao, Haiyu Yi, Shuai Li, Hongsheng |
| contents | Dot-product attention has wide applications in computer vision and natural language processing. However, its memory and computational costs grow quadratically with the input size. Such growth prohibits its application on high-resolution inputs. To remedy this drawback, this paper proposes a novel efficient attention mechanism equivalent to dot-product attention but with substantially less memory and computational costs. Its resource efficiency allows more widespread and flexible integration of attention modules into a network, which leads to better accuracies. Empirical evaluations demonstrated the effectiveness of its advantages. Efficient attention modules brought significant performance boosts to object detectors and instance segmenters on MS-COCO 2017. Further, the resource efficiency democratizes attention to complex models, where high costs prohibit the use of dot-product attention. As an exemplar, a model with efficient attention achieved state-of-the-art accuracies for stereo depth estimation on the Scene Flow dataset. Code is available at https://github.com/cmsflash/efficient-attention. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_1812_01243 |
| institution | arXiv |
| publishDate | 2018 |
| record_format | arxiv |
| spellingShingle | Efficient Attention: Attention with Linear Complexities Shen, Zhuoran Zhang, Mingyuan Zhao, Haiyu Yi, Shuai Li, Hongsheng Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning I.2.6; I.2.10; I.4.8; I.4.6 Dot-product attention has wide applications in computer vision and natural language processing. However, its memory and computational costs grow quadratically with the input size. Such growth prohibits its application on high-resolution inputs. To remedy this drawback, this paper proposes a novel efficient attention mechanism equivalent to dot-product attention but with substantially less memory and computational costs. Its resource efficiency allows more widespread and flexible integration of attention modules into a network, which leads to better accuracies. Empirical evaluations demonstrated the effectiveness of its advantages. Efficient attention modules brought significant performance boosts to object detectors and instance segmenters on MS-COCO 2017. Further, the resource efficiency democratizes attention to complex models, where high costs prohibit the use of dot-product attention. As an exemplar, a model with efficient attention achieved state-of-the-art accuracies for stereo depth estimation on the Scene Flow dataset. Code is available at https://github.com/cmsflash/efficient-attention. |
| title | Efficient Attention: Attention with Linear Complexities |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning I.2.6; I.2.10; I.4.8; I.4.6 |
| url | https://arxiv.org/abs/1812.01243 |