arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2104.09497cs.CV
收藏

图像超分辨率的注意力网络中的注意力

Attention in Attention Network for Image Super-Resolution

Haoyu Chen, Jinjin Gu, Zhi Zhang

摘要:

卷积神经网络在过去十年中在单图像超分辨率(SISR)方面取得了显着进展。在SISR的最新进展中,注意力机制是高性能SR模型的关键。然而,注意力机制在SISR中为什么以及如何起作用仍然不清楚。在这项工作中,我们试图量化和可视化的注意力机制在SISR和显示,并不是所有的注意力模块是同样有益的。然后,我们提出了注意力网络中的注意力(A $^2$ N)更有效和准确的SISR。具体地,A $^2$ N由非注意分支和耦合注意分支组成。提出了一个动态注意力模块来生成这两个分支的权重,以动态地抑制不必要的注意力调整,其中权重根据输入特征自适应地变化。这使得注意力模块能够专门针对有益的示例而没有其他惩罚,从而大大提高了注意力网络的容量,并且参数开销很少。实验结果表明,我们最终的模型A $^2$ N可以实现优越的权衡性能相比,国家的最先进的网络类似的大小。代码可在https://github.com/haoyuc/A2N上获得。

英文摘要:

Convolutional neural networks have allowed remarkable advances in single image super-resolution (SISR) over the last decade. Among recent advances in SISR, attention mechanisms are crucial for high-performance SR models. However, the attention mechanism remains unclear on why and how it works in SISR. In this work, we attempt to quantify and visualize attention mechanisms in SISR and show that not all attention modules are equally beneficial. We then propose attention in attention network (A$^2$N) for more efficient and accurate SISR. Specifically, A$^2$N consists of a non-attention branch and a coupling attention branch. A dynamic attention module is proposed to generate weights for these two branches to suppress unwanted attention adjustments dynamically, where the weights change adaptively according to the input features. This allows attention modules to specialize to beneficial examples without otherwise penalties and thus greatly improve the capacity of the attention network with few parameters overhead. Experimental results demonstrate that our final model A$^2$N could achieve superior trade-off performances comparing with state-of-the-art networks of similar sizes. Codes are available at https://github.com/haoyuc/A2N.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑