Skip to content

Latest commit

 

History

History
91 lines (57 loc) · 2.08 KB

pretrained_resnet.md

File metadata and controls

91 lines (57 loc) · 2.08 KB
Raw

Note: This is a generated markdown export from the Jupyter notebook file pretrained_resnet.ipynb. You can also view the notebook with the nbviewer from Jupyter.

Pre-trained model ResNet

%matplotlib inline

import urllib
from PIL import Image

from tensorflow.keras.applications.resnet50 import ResNet50
from tensorflow.keras.preprocessing import image
from tensorflow.keras.applications.resnet50 import preprocess_input, decode_predictions

import numpy as np
import matplotlib.pyplot as plt
bears = [
    'https://upload.wikimedia.org/wikipedia/commons/3/3d/Polar_Bear_AdF.jpg',
    'https://upload.wikimedia.org/wikipedia/commons/7/7d/Brown_bear.jpg',
    'https://upload.wikimedia.org/wikipedia/commons/6/6e/Canadian_Rockies_-_the_bear_at_Lake_Louise.jpg',
    'https://upload.wikimedia.org/wikipedia/commons/9/90/Koala_in_Zoo_Duisburg.jpg'
]

imgs = []

fig, axes = plt.subplots(nrows=1, ncols=len(bears))

for index, b in enumerate(bears):
    img = Image.open(urllib.request.urlopen(b))
    img = img.resize(size=(224, 224), resample=Image.NEAREST)
    _ = axes[index].imshow(img)
    imgs.append(img)

png

X = []
for i in imgs:
    x = image.img_to_array(i)
    x = np.expand_dims(x, axis=0)
    x = preprocess_input(x)
    X.append(x)

X = np.reshape(X, (4, 224, 224, 3))
print(np.shape(X))
(4, 224, 224, 3)

model = ResNet50(weights='imagenet')
preds = model.predict(X)
fig, axes = plt.subplots(nrows=len(bears), ncols=1, figsize=(10,40))

for index, predictions in enumerate(decode_predictions(preds, top=3)):
    
    axes[index].imshow(imgs[index])
   
    title = ''
    for p in predictions:
        tag = p[1]
        prob = p[2]
        title += '{} with p: {:.5f}
'.format(tag, prob)
     
    axes[index].set_title(title)

png