from __future__ import print_function
from keras.optimizers import SGD, RMSprop
from cnn_functions import rate_scheduler, train_model_sample
from model_zoo import feature_net_61x61 as the_model
import os
import datetime
import numpy as np
batch_size = 256
n_classes = 3
n_epoch = 25
model = the_model(n_channels=2, n_features=3, reg=1e-4, drop=0.5)
dataset = "nuclei_all_61x61"
direc_save = "/home/nquach/DeepCell2/trained_networks/"
direc_data = "/home/nquach/DeepCell2/training_data_npz/"
optimizer = RMSprop(lr=0.001, rho=0.95, epsilon=1e-8)
lr_sched = rate_scheduler(lr=0.001, decay=0.95)
expt = "feature_net_61x61_drop_reg4"
iterate = 4
train_model_sample(model=model,
dataset=dataset,
optimizer=optimizer,
expt=expt,
it=iterate,
batch_size=batch_size,
n_epoch=n_epoch,
from __future__ import print_function
from keras.optimizers import SGD, RMSprop
from cnn_functions import rate_scheduler, train_model_sample
from model_zoo import feature_net_61x61 as the_model
import os
import datetime
import numpy as np
batch_size = 256
n_classes = 3
n_epoch = 25
model = the_model(n_channels=2, n_features=3, reg=0, drop=0)
dataset = "HeLa_all_61x61"
direc_save = "/home/nquach/DeepCell2/trained_networks/"
direc_data = "/home/nquach/DeepCell2/training_data_npz/"
optimizer = RMSprop(lr=0.001, rho=0.95, epsilon=1e-8)
lr_sched = rate_scheduler(lr=0.001, decay=0.95)
expt = "feature_net_61x61_reg0"
iterate = 0
train_model_sample(model=model,
dataset=dataset,
optimizer=optimizer,
expt=expt,
it=iterate,
batch_size=batch_size,
n_epoch=n_epoch,
from __future__ import print_function
from keras.optimizers import SGD, RMSprop
from cnn_functions import rate_scheduler, train_model_sample
from model_zoo import bn_feature_net_61x61 as the_model
import os
import datetime
import numpy as np
batch_size = 256
n_classes = 3
n_epoch = 25
model = the_model(n_channels=2, n_features=3, reg=1e-5)
dataset = "HeLa_set3_61x61"
direc_save = "/home/nquach/DeepCell2/trained_networks/"
direc_data = "/home/nquach/DeepCell2/training_data_npz/"
optimizer = SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)
lr_sched = rate_scheduler(lr=0.01, decay=0.95)
expt = "bn_61x61_HeLa_set3"
iterate = 1
train_model_sample(model=model,
dataset=dataset,
optimizer=optimizer,
expt=expt,
it=iterate,
batch_size=batch_size,
n_epoch=n_epoch,
data_location = os.path.join(direc_name, 'RawImages')
output_location = os.path.join(direc_name, 'Output')
channel_names = ['Phase', 'Far-red']
win_x = 30
win_y = 30
image_size_x, image_size_y = get_image_sizes(data_location, channel_names)
image_size_x /= 2
image_size_y /= 2
"""
Define model
"""
trained_network_directory = "/home/vanvalen/DeepCell2/trained_networks/"
file_name_save = os.path.join(trained_network_directory,
"2016-07-12_HeLa_all_61x61_bn_shear_0.h5")
model = the_model(batch_input_shape=(1, 2, image_size_x + win_x,
image_size_y + win_x),
weights_path=file_name_save)
"""
Run model on directory
"""
run_model_on_directory(data_location,
channel_names,
output_location,
model=model,
win_x=win_x,
win_y=win_y,
std=False)