Stand-Alone Self-Attention in Vision Models
NeurIPS, pp. 68-80, 2019.
Convolutions are a fundamental building block of modern computer vision systems. Recent approaches have argued for going beyond convolutions in order to capture long-range dependencies. These efforts focus on augmenting convolutional models with content-based interactions, such as self-attention and non-local means, to achieve gains on a ...More
PPT (Upload PPT)