[英]Visualize the learned filter of each CNN layer
anyone please tell me how to visualize the learned filter of each CNN layer? 有人请告诉我如何可视化每个CNN层的学习过滤器? The following answers tell me how to only visualize the learned filters of the first CNN layer, but could not visulize the other CNN layers.
以下答案告诉我如何仅可视化第一个CNN层的学习过滤器,而不能可视化其他CNN层。
1) You can just recover the filters and use Matlab's functions to display them as images. 1)您可以恢复过滤器并使用Matlab的功能将其显示为图像。 For example after loading a pretrained net from http://www.vlfeat.org/matconvnet/pretrained/ :
例如,从http://www.vlfeat.org/matconvnet/pretrained/加载预训练的网络后:
imshow( net.layers{1}.filters(:, :, 3, 1), [] ) ;
2) You may find the VLFeat function vl_imarraysc useful to display several filters. 2)您可能会发现VLFeat函数vl_imarraysc对显示多个过滤器很有用。 http://www.vlfeat.org/matlab/vl_imarraysc.html
http://www.vlfeat.org/matlab/vl_imarraysc.html
For visualizing filters in intermediate layers. 用于可视化中间层中的过滤器。 There are several techniques:
有几种技术:
(1) show one or three channels as grayscale or RGB at a time. (1)一次显示一个或三个通道为灰度或RGB。 It's not very informative since they filters of ResNet and VGG are small 3x3.
由于它们的ResNet和VGG过滤器均为3x3小型,因此它的信息不是很丰富。
(2) Turn off other units. (2)关闭其他单元。 Backpropgate only this unit to the input space.
仅将本机反向传播到输入空间。 You can see a pattern that reflects what this unit cares about.
您可以看到一种模式,该模式反映了本机的要求。 There are many papers that use similar techniques.
有许多使用类似技术的论文。 eg, Zeiler, Matthew D., and Rob Fergus.
例如Zeiler,Matthew D.和Rob Fergus。 "Visualizing and understanding convolutional networks."
“可视化和理解卷积网络。” European conference on computer vision.
欧洲计算机视觉会议。 2014.
2014。
(3) Find input patches that maximally activate this unit and see what they are. (3)找到可以最大程度激活本机的输入音色,并查看它们是什么。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.