這幾天看官方提供基於iPython notebook的學習手冊,提供了一些基於Python程式碼的呼叫例項.

這一篇是關於學習特徵視覺化的.地址如下:

http://nbviewer.ipython.org/github/BVLC/caffe/blob/master/examples/filter_visualization.ipynb

其中有幾個地方程式碼操作有點煩人.,原文描述如下

The second layer filters, conv2

There are 256 filters, each of which has dimension 5 x 5 x 48. We show only the first 48 filters, with each channel shown separately, so that each filter is a row.

filters = net.params['conv2'][0].data
vis_square(filters[:48].reshape(48**2, 5, 5))

只顯示了48個卷積核的引數資訊.怎麼才能顯示全部的呢,  呼叫 filters.size.得到307200這麼一個數值,這個是256*48*5*5的值,就是 有256個卷積核,每個是5*5大小,有48組,剛好對應要求. 於是這麼改一下就能看到全部:

                           vis_square(filters[:256].reshape(48*256, 5,
 
 5))

:))