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
The following image classification models (with weights trained on ImageNet) are available:
All of these architectures (except Xception and MobileNet) are compatible with both TensorFlow and Theano, and 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 TensorFlow data format convention, “Width-Height-Depth”.
The Xception model is only available for TensorFlow, due to its reliance on
SeparableConvolution layers. The MobileNet model is only available for TensorFlow, due to its reliance on
# instantiate the model model <- application_resnet50(weights = 'imagenet') # load the image img_path <- "elephant.jpg" img <- image_load(img_path, target_size = c(224,224)) x <- image_to_array(img) # ensure we have a 4d tensor with single element in the batch dimension, # the preprocess the input for prediction using resnet50 x <- array_reshape(x, c(1, dim(x))) x <- imagenet_preprocess_input(x) # make predictions then decode and print them preds <- model %>% predict(x) imagenet_decode_predictions(preds, top = 3)[]
class_name class_description score 1 n02504013 Indian_elephant 0.90117526 2 n01871265 tusker 0.08774310 3 n02504458 African_elephant 0.01046011
base_model <- application_vgg19(weights = 'imagenet') model <- keras_model(inputs = base_model$input, outputs = get_layer(base_model, 'block4_pool')$output) img_path <- "elephant.jpg" img <- image_load(img_path, target_size = c(224,224)) x <- image_to_array(img) x <- array_reshape(x, c(1, dim(x))) x <- imagenet_preprocess_input(x) block4_pool_features <- model %>% predict(x)
# create the base pre-trained model base_model <- application_inception_v3(weights = 'imagenet', include_top = FALSE) # add our custom layers predictions <- base_model$output %>% layer_global_average_pooling_2d() %>% layer_dense(units = 1024, activation = 'relu') %>% layer_dense(units = 200, activation = 'softmax') # this is the model we will train model <- keras_model(inputs = base_model$input, outputs = predictions) # first: train only the top layers (which were randomly initialized) # i.e. freeze all convolutional InceptionV3 layers freeze_weights(base_model) # compile the model (should be done *after* setting layers to non-trainable) model %>% compile(optimizer = 'rmsprop', loss = 'categorical_crossentropy') # train the model on the new data for a few epochs model %>% fit_generator(...) # at this point, the top layers are well trained and we can start fine-tuning # convolutional layers from inception V3. We will freeze the bottom N layers # and train the remaining top layers. # let's visualize layer names and layer indices to see how many layers # we should freeze: layers <- base_model$layers for (i in 1:length(layers)) cat(i, layers[[i]]$name, "\n") # we chose to train the top 2 inception blocks, i.e. we will freeze # the first 172 layers and unfreeze the rest: freeze_weights(base_model, from = 1, to = 172) unfreeze_weights(base_model, from = 173) # we need to recompile the model for these modifications to take effect # we use SGD with a low learning rate model %>% compile( optimizer = optimizer_sgd(lr = 0.0001, momentum = 0.9), loss = 'categorical_crossentropy' ) # we train our model again (this time fine-tuning the top 2 inception blocks # alongside the top Dense layers model %>% fit_generator(...)
The VGG16 model is the basis for the Deep dream Keras example script.