qubvel / resnet_152

Pretrained on ImageNet ResNet-152 model in Keras

ResNet-152 model for Keras

Reference:

• Deep Residual Learning for Image Recognition

Extention of model taken here.
Model made in Keras style with pretrained weights from ImageNet provided in release (load automatically during model initialization).

Example

Inference example for this picture:

Code

import numpy as np
from skimage.io import imread
from skimage.transform import resize
from keras.applications.imagenet_utils import decode_predictions
from keras.applications.imagenet_utils import preprocess_input
from model import ResNet152

# create model
model = ResNet152()

# define function for input preprocessing
def preprocess(x):
    x = resize(x, (224,224), mode='constant') * 255
    x = preprocess_input(x)
    if x.ndim == 3:
        x = np.expand_dims(x, 0)
    return x

# prepare image
img = imread('./imgs/cat.jpg')
x = preprocess(img)

# make prediction and decode it
y = model.predict(x)
pred_title = decode_predictions(y, top=1)[0][0][1]

# print result
print(pred_title)
### tiget_cat