论文标题
注意特征融合
Attentional Feature Fusion
论文作者
论文摘要
特征融合（来自不同层或分支的特征的组合）是现代网络体系结构的无所不在的一部分。它通常是通过简单操作（例如求和或串联）实现的，但这可能不是最佳选择。在这项工作中，我们提出了一个统一和一般的方案，即注意特征融合，该方案适用于大多数常见的情况，包括由短和长跳连接以及在开始层内引起的特征融合。为了更好地融合不一致的语义和尺度的功能，我们提出了一个多尺度的通道注意模块，该模块解决了在不同尺度上给出的融合功能时出现的问题。我们还证明，特征地图的初始整合可能会成为瓶颈，并且可以通过增加另一个关注度来缓解此问题，我们将其称为迭代的注意特征融合。在CIFAR-100和Imagenet数据集上，我们的模型较少的层或参数均优于最先进的网络，这表明与直接对应物相比，功能融合的更复杂的注意力机制具有巨大的潜力，可以始终如一地产生更好的结果。我们的代码和训练有素的模型可在线提供。
Feature fusion, the combination of features from different layers or branches, is an omnipresent part of modern network architectures. It is often implemented via simple operations, such as summation or concatenation, but this might not be the best choice. In this work, we propose a uniform and general scheme, namely attentional feature fusion, which is applicable for most common scenarios, including feature fusion induced by short and long skip connections as well as within Inception layers. To better fuse features of inconsistent semantics and scales, we propose a multi-scale channel attention module, which addresses issues that arise when fusing features given at different scales. We also demonstrate that the initial integration of feature maps can become a bottleneck and that this issue can be alleviated by adding another level of attention, which we refer to as iterative attentional feature fusion. With fewer layers or parameters, our models outperform state-of-the-art networks on both CIFAR-100 and ImageNet datasets, which suggests that more sophisticated attention mechanisms for feature fusion hold great potential to consistently yield better results compared to their direct counterparts. Our codes and trained models are available online.
