文章作者：Tyan
博客：noahsnail.com ?|? CSDN ?|? 簡書

本文主要是對Caffe中mnist數據集上訓練的LeNet模型進行結構分析和可視化。

import caffe
import numpy as np
import matplotlib.pyplot as plt

# 定義LeNet模型信息
deploy = 'lenet.prototxt'
model = 'lenet_iter_10000.caffemodel'

# 加載模型
net = caffe.Net(deploy, model, caffe.TEST)

# 計算均值
# blob = caffe.proto.caffe_pb2.BlobProto()
# bin_mean = open(mean_file, 'rb' ).read()
# blob.ParseFromString(bin_mean)
# arr = np.array(caffe.io.blobproto_to_array(blob))
# npy_mean = arr[0]
# mu = npy_mean.mean(1).mean(1)

# init transformer
transformer = caffe.io.Transformer({'data': net.blobs['data'].data.shape})
transformer.set_transpose('data', (2, 0, 1))
# transformer.set_mean('data', mu)
transformer.set_raw_scale('data', 255)
# transformer.set_channel_swap('data', (2, 1, 0))

# get certain layer feature
def init(pimg, lay_name):
    global transformer
    global net
    image = caffe.io.load_image(pimg, color = False)
    image
    transformed_image = transformer.preprocess('data', image)
    net.blobs['data'].data[...] = transformed_image
    output = net.forward()
    result = output[lay_name]
    return result

# Test
result = init('test.jpg', 'prob')
print result.shape
print result

(1, 10)
[[ 1.  0.  0.  0.  0.  0.  0.  0.  0.  0.]]

LeNet網絡的所有layer以及layer的輸出數據
data: 輸入圖片數據大小為28*28
conv1: 20個卷積核，卷積之后feature map大小24*24
pool1: pooling后feature map大小變為12*12, 共20層
conv2: 50個卷積核, 卷積之后feature map大小為8*8
pool2: pooling后feature map大小變為4*4, 共50層
ip1: 全連接層一, 500個結點
ip2: 全連接層二, 10個結點
prob: 對ip2進行softmax

備注: conv1之后得到20個feature map, conv2有50個卷積核, 每個卷積核在20個feature map卷積之后, 20個卷積之后的feature map對應位置上的點的數據累加之后取激活函數(ReLU)得到該卷積核的對應的feature map, 因此conv2執行之后的feature map個數為50, 而不是50*20.

# all layer name and blob shape
# blob shape is (batch_size, channel_dim, height, width).
for layer_name, blob in net.blobs.iteritems():
    print layer_name + '\t' + str(blob.data.shape)

data  (1, 1, 28, 28)
conv1 (1, 20, 24, 24)
pool1 (1, 20, 12, 12)
conv2 (1, 50, 8, 8)
pool2 (1, 50, 4, 4)
ip1 (1, 500)
ip2 (1, 10)
prob  (1, 10)

LeNet網絡的權重(weights + biases)
conv1: 20個卷積核, weights大小為5*5, 20個biases
conv2: 50個卷積核, weights大小為5*5, 50個biases
ip1: conv2之后得到50個4*4大小的feature map, 排列起來大小為800, 與ip1的500個結點進行全連接, weights個數為500*800, biases個數為500
ip2: ip1的500個結點與ip2的10個結點進行全連接, weights個數為500*10, biases個數為10

# all layer name and parameters shape
# param[0] is weights, param[1] is biases
# weights shape is (output_channels, input_channels, filter_height, filter_width)
# biases shape is (output_channels,)
for layer_name, param in net.params.iteritems():
    print layer_name + '\t' + str(param[0].data.shape) + '\t' + str(param[1].data.shape)

conv1 (20, 1, 5, 5) (20,)
conv2 (50, 20, 5, 5)  (50,)
ip1 (500, 800)  (500,)
ip2 (10, 500) (10,)

numpy pad
padding分為四部分
第一部分: (0, n ** 2 - data.shape[0]), 補充方陣的缺少的部分, 0表示前面不補, 后面補n ** 2 - data.shape[0]列
第二部分: (0, 1)表示每個filter的前面不補, 后面補1列, filter補了一行
第三部分: (0, 1)表示每個filter的前面不補, 后面補1列, filter補了一列
第四部分: (0, 0)剩下的不補充數據

# param(weights) visualization
def visualization(data):
    # normalize data for display
    data = (data - data.min()) / (data.max() - data.min())
    
    # force the number of filters to be square
    n = int(np.ceil(np.sqrt(data.shape[0])))
    
    # add some space between filters
    padding = (((0, n ** 2 - data.shape[0]), (0, 1), (0, 1)) + ((0, 0),) * (data.ndim - 3)) 
    data = np.pad(data, padding, mode = 'constant', constant_values = 1)
    
    # tile the filters into an image
    data = data.reshape((n, n) + data.shape[1:]).transpose((0, 2, 1, 3) + tuple(range(4, data.ndim + 1)))
    data = data.reshape((n * data.shape[1], n * data.shape[3]) + data.shape[4:])
    plt.imshow(data, cmap='gray')
    plt.axis('off')
    plt.show()

# feature map visualization
feature_map = net.blobs['conv1'].data[0]
visualization(feature_map)

png

# filter visualization
filters = net.params['conv1'][0].data
visualization(filters.reshape(20, 5, 5))

png