论文笔记-Squeeze-and-Excitation Networks

作者提出为了增强网络的表达能力，现有的工作显示了加强空间编码的作用。在这篇论文里面，作者重点关注channel上的信息，提出了“Squeeze-and-Excitation"（SE）block，实际上就是显式的让网络关注channel之间的信息 （adaptively recalibrates channel-wise feature responsesby explicitly modelling interdependencies between channels.）。SEnets取得了ILSVRC2017的第一名, top-5 error 2.251%

之前的一些架构设计关注空间依赖
Inception architectures: embedding multi-scale processes in its modules
Resnet, stack hourglass
spatial attention: Spatial transformer networks

作者的设计思路：
we investigate a different
aspect of architectural design - the channel relationship
Our goal is to improve the representational power of a network by explicitly
modelling the interdependencies between the channels of its
convolutional features. To achieve this, we propose a mechanism that allows the network to perform feature recalibration, through which it can learn to use global information
to selectively emphasise informative features and suppress
less useful ones. 作者希望能够对卷积特征进行recalibration，根据后文我的理解就是对channel进行加权了。

相关工作
网络结构：
VGGNets, Inception models, BN, Resnet, Densenet, Dual path network
其他方式：Grouped convolution, Multi-branch convolution, Cross-channel correlations
This approach reflects an assumption that channel relationships can
be formulated as a composition of instance-agnostic functions with local receptive fields.
Attention, gating mechanisms

SE block

({F_{tr}}:X in R{^{W' imes H' imes C'}},{kern 1pt} {kern 1pt} {kern 1pt} {kern 1pt} {kern 1pt} U in {kern 1pt} {kern 1pt} {R^{W imes H imes C}})
设(V = [v_1, v_2, ..., v_C])表示学习到的filter kernel, (v_c)表示第c个filter的参数，那么(F_{tr})的输出(U = [u_1,u_2,...,u_C]):

[{u_c} = { m{ }}{{ m{v}}_c} * X = sumlimits_{s = 1}^{C'} {v_c^s} * {x^s} ]

(v_c^s)是一个channel的kernel，一个新产生的channel是原有所有channel与相应的filter kernel卷积的和。channel间的关系隐式的包含在(v_c)中，但是这些信息和空间相关性纠缠在一起了，作者的目标就是让网络更加关注有用的信息。分成了Squeeze和Excitation两步来完成目的。
Squeeze
现有网络的问题：由于卷积实在local receptive field做的，因此每个卷积单元只能关注这个field内的空间信息。
为了减轻这个问题，提出了Squeeze操作将全局的空间信息编码到channel descriptor中，具体而言是通过global average pooling操作完成的。

[{z_c} = {F_{sq}}({u_c}) = {1 over {W imes H}}sumlimits_{i = 1}^W {sumlimits_{j = 1}^H {{u_c}(i,j)} } ]

就是求每个channel的均值，作为全局的描述。
Excitation: Adaptive Recalibration
为了利用Squeeze得到的信息，提出了第二个op,这个op需要满足2个要求：一个是足够灵活，需要能够学习channel间的非线性关系，另一个就是能够学习non-mutually-exclusive关系，这个词我的理解是非独占性，可能是说多个channnel之间会有各种各样的关系吧。

[s = {F_{ex}}(z,W) = sigma (g(z,W)) = sigma ({W_2}delta ({W_1}z)) ]

$delta (是ReLu,){W_1} in {R^{{C over r} imes C}}(，){W_2} in {R^{C imes {C over r}}}(，)W_1(是bottleneck，降低channel数，)W_2(是增加channel数，)gamma(设置为16。最终再将)U(用)s$来scale，其实也就是加权了。这样就得到了一个block的输出。

[{x_c} = {F_{scale}}({u_c},{s_c}) = {s_c} cdot {u_c} ]

(F_{scale})表示feature map (u_c in R^{W imes H})和(s_c)的channel-wise乘法

The activations act as channel weights
adapted to the input-specific descriptor z. In this regard,
SE blocks intrinsically introduce dynamics conditioned on
the input, helping to boost feature discriminability

3. Example
   
   SE block可以很方便的加到其他网络结构上。
4. Mxnet code

squeeze = mx.sym.Pooling(data=bn3, global_pool=True, kernel=(7, 7), pool_type='avg', name=name + '_squeeze')
squeeze = mx.symbol.Flatten(data=squeeze, name=name + '_flatten')
excitation = mx.symbol.FullyConnected(data=squeeze, num_hidden=int(num_filter*ratio), name=name + '_excitation1')#bottleneck
excitation = mx.sym.Activation(data=excitation, act_type='relu', name=name + '_excitation1_relu')
excitation = mx.symbol.FullyConnected(data=excitation, num_hidden=num_filter, name=name + '_excitation2')
excitation = mx.sym.Activation(data=excitation, act_type='sigmoid', name=name + '_excitation2_sigmoid')
bn3 = mx.symbol.broadcast_mul(bn3, mx.symbol.reshape(data=excitation, shape=(-1, num_filter, 1, 1)))

5. 网络结构
   

6. Experiments
   

参考文献：
[1] Hu, Jie, Li Shen, and Gang Sun. "Squeeze-and-excitation networks." arXiv preprint arXiv:1709.01507 (2017).

欢迎关注公众号：vision_home 共同学习，不定期分享论文和资源