Keras Applications

원본 링크 : https://keras.io/api/applications/
최종 확인 : 2024-11-25

Keras Applications are deep learning models that are made available alongside pre-trained weights. These models can be used for prediction, feature extraction, and fine-tuning.

Weights are downloaded automatically when instantiating a model. They are stored at ~/.keras/models/.

Upon instantiation, the models will be built according to the image data format set in your Keras configuration file at ~/.keras/keras.json. For instance, if you have set image_data_format=channels_last, then any model loaded from this repository will get built according to the data format convention “Height-Width-Depth”.

Available models

Model	Size (MB)	Top-1 Accuracy	Top-5 Accuracy	Parameters	Depth	Time (ms) per inference step (CPU)	Time (ms) per inference step (GPU)
Xception	88	79.0%	94.5%	22.9M	81	109.4	8.1
VGG16	528	71.3%	90.1%	138.4M	16	69.5	4.2
VGG19	549	71.3%	90.0%	143.7M	19	84.8	4.4
ResNet50	98	74.9%	92.1%	25.6M	107	58.2	4.6
ResNet50V2	98	76.0%	93.0%	25.6M	103	45.6	4.4
ResNet101	171	76.4%	92.8%	44.7M	209	89.6	5.2
ResNet101V2	171	77.2%	93.8%	44.7M	205	72.7	5.4
ResNet152	232	76.6%	93.1%	60.4M	311	127.4	6.5
ResNet152V2	232	78.0%	94.2%	60.4M	307	107.5	6.6
InceptionV3	92	77.9%	93.7%	23.9M	189	42.2	6.9
InceptionResNetV2	215	80.3%	95.3%	55.9M	449	130.2	10.0
MobileNet	16	70.4%	89.5%	4.3M	55	22.6	3.4
MobileNetV2	14	71.3%	90.1%	3.5M	105	25.9	3.8
DenseNet121	33	75.0%	92.3%	8.1M	242	77.1	5.4
DenseNet169	57	76.2%	93.2%	14.3M	338	96.4	6.3
DenseNet201	80	77.3%	93.6%	20.2M	402	127.2	6.7
NASNetMobile	23	74.4%	91.9%	5.3M	389	27.0	6.7
NASNetLarge	343	82.5%	96.0%	88.9M	533	344.5	20.0
EfficientNetB0	29	77.1%	93.3%	5.3M	132	46.0	4.9
EfficientNetB1	31	79.1%	94.4%	7.9M	186	60.2	5.6
EfficientNetB2	36	80.1%	94.9%	9.2M	186	80.8	6.5
EfficientNetB3	48	81.6%	95.7%	12.3M	210	140.0	8.8
EfficientNetB4	75	82.9%	96.4%	19.5M	258	308.3	15.1
EfficientNetB5	118	83.6%	96.7%	30.6M	312	579.2	25.3
EfficientNetB6	166	84.0%	96.8%	43.3M	360	958.1	40.4
EfficientNetB7	256	84.3%	97.0%	66.7M	438	1578.9	61.6
EfficientNetV2B0	29	78.7%	94.3%	7.2M	-	-	-
EfficientNetV2B1	34	79.8%	95.0%	8.2M	-	-	-
EfficientNetV2B2	42	80.5%	95.1%	10.2M	-	-	-
EfficientNetV2B3	59	82.0%	95.8%	14.5M	-	-	-
EfficientNetV2S	88	83.9%	96.7%	21.6M	-	-	-
EfficientNetV2M	220	85.3%	97.4%	54.4M	-	-	-
EfficientNetV2L	479	85.7%	97.5%	119.0M	-	-	-
ConvNeXtTiny	109.42	81.3%	-	28.6M	-	-	-
ConvNeXtSmall	192.29	82.3%	-	50.2M	-	-	-
ConvNeXtBase	338.58	85.3%	-	88.5M	-	-	-
ConvNeXtLarge	755.07	86.3%	-	197.7M	-	-	-
ConvNeXtXLarge	1310	86.7%	-	350.1M	-	-	-

The top-1 and top-5 accuracy refers to the model’s performance on the ImageNet validation dataset.

Depth refers to the topological depth of the network. This includes activation layers, batch normalization layers etc.

Time per inference step is the average of 30 batches and 10 repetitions.

CPU: AMD EPYC Processor (with IBPB) (92 core)
RAM: 1.7T
GPU: Tesla A100
Batch size: 32

Depth counts the number of layers with parameters.

Usage examples for image classification models

Classify ImageNet classes with ResNet50

import keras
from keras.applications.resnet50 import ResNet50
from keras.applications.resnet50 import preprocess_input, decode_predictions
import numpy as np
model = ResNet50(weights='imagenet')
img_path = 'elephant.jpg'
img = keras.utils.load_img(img_path, target_size=(224, 224))
x = keras.utils.img_to_array(img)
x = np.expand_dims(x, axis=0)
x = preprocess_input(x)
preds = model.predict(x)
print('Predicted:', decode_predictions(preds, top=3)[0])

Extract features with VGG16

import keras
from keras.applications.vgg16 import VGG16
from keras.applications.vgg16 import preprocess_input
import numpy as np
model = VGG16(weights='imagenet', include_top=False)
img_path = 'elephant.jpg'
img = keras.utils.load_img(img_path, target_size=(224, 224))
x = keras.utils.img_to_array(img)
x = np.expand_dims(x, axis=0)
x = preprocess_input(x)
features = model.predict(x)

Extract features from an arbitrary intermediate layer with VGG19

from keras.applications.vgg19 import VGG19
from keras.applications.vgg19 import preprocess_input
from keras.models import Model
import numpy as np
base_model = VGG19(weights='imagenet')
model = Model(inputs=base_model.input, outputs=base_model.get_layer('block4_pool').output)
img_path = 'elephant.jpg'
img = keras.utils.load_img(img_path, target_size=(224, 224))
x = keras.utils.img_to_array(img)
x = np.expand_dims(x, axis=0)
x = preprocess_input(x)
block4_pool_features = model.predict(x)

Fine-tune InceptionV3 on a new set of classes

from keras.applications.inception_v3 import InceptionV3
from keras.models import Model
from keras.layers import Dense, GlobalAveragePooling2D
base_model = InceptionV3(weights='imagenet', include_top=False)
x = base_model.output
x = GlobalAveragePooling2D()(x)
x = Dense(1024, activation='relu')(x)
predictions = Dense(200, activation='softmax')(x)
model = Model(inputs=base_model.input, outputs=predictions)
for layer in base_model.layers:
    layer.trainable = False
model.compile(optimizer='rmsprop', loss='categorical_crossentropy')
model.fit(...)
for i, layer in enumerate(base_model.layers):
   print(i, layer.name)
for layer in model.layers[:249]:
   layer.trainable = False
for layer in model.layers[249:]:
   layer.trainable = True
from keras.optimizers import SGD
model.compile(optimizer=SGD(lr=0.0001, momentum=0.9), loss='categorical_crossentropy')
model.fit(...)

Build InceptionV3 over a custom input tensor

from keras.applications.inception_v3 import InceptionV3
from keras.layers import Input
input_tensor = Input(shape=(224, 224, 3))
model = InceptionV3(input_tensor=input_tensor, weights='imagenet', include_top=True)