Filter Visualization_Keras_VGG16.py

# -*- coding: utf-8 -*-
"""
Created on Wed Jan  1 20:11:42 2020

@author: liuji
"""


from keras.applications import VGG16
from keras import backend as K
import matplotlib.pyplot as plt

model = VGG16(weights='imagenet',include_top = False)


import numpy as np


def deprocess_image(x):
    x -= x.mean()
    x /= (x.std() + 1e-5)
    x *= 0.1
    
    x += 0.5
    x = np.clip(x,0,1)

    x *= 255
    x = np.clip(x,0,255).astype('uint8')
    return x

def generate_pattern(layer_name, filter_index, size =150):
    layer_output = model.get_layer(layer_name).output
    loss = K.mean(layer_output[:,:,:,filter_index])

    grads = K.gradients(loss,model.input)[0]

    grads /= (K.sqrt(K.mean(K.square(grads))) + 1e-5)

    iterate = K.function([model.input],[loss,grads])

    input_img_data = np.random.random((1,size,size,3)) * 20 + 128.

    step = 1.
    for i in range(40):
        loss_value,grads_value = iterate([input_img_data])
        input_img_data += grads_value * step

    img = input_img_data[0]
    return deprocess_image(img)

# plt.imshow(generate_pattern('block1_conv1',0))
# plt.show()

layer_name = 'block3_conv1'
size = 64 
margin = 5

results = np.zeros((8 * size + 7 * margin, 8 * size + 7 * margin,3))

for i in range(8):
    for j in range(8):
        filter_img = generate_pattern(layer_name,i + (j * 8),size = size)
        horizontal_start = i * size + i * margin
        horizontal_end = horizontal_start + size
        vertical_start = j * size + j * margin
        vertical_end = vertical_start + size
        results[horizontal_start:horizontal_end,vertical_start:vertical_end,:] = filter_img

results = np.clip(results,0,255).astype('uint8')

plt.figure(figsize=(20,20))
plt.imshow(results)
plt.show()