merged.py

#!/usr/local/bin/python3
import numpy as np
np.random.seed(1337)  # for reproducibility
from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Merge
import matplotlib.pyplot as plt
import pickle
import json
import mfcc_model
import tempotrack_model
import spectral_contrast_peaks_model
import theano as T
from os.path import exists
import os


batch_size = 50
nb_epoch = 50

numGenres = 3

print("creating model")
# create model
y = pickle.load(open("pickled_vectors/mfcc_coefficients_label.pickle","rb"))
y_test = pickle.load(open("pickled_vectors/mfcc_coefficients_evaluation_label.pickle","rb"))
# X_1 = pickle.load(open("pickled_vectors/tempotracker_tempo_training_vector.pickle","rb"))
# X_test_1 = pickle.load(open("pickled_vectors/tempotracker_tempo_evaluation_training_vector.pickle","rb"))

# model_1 = tempotrack_model.tempotrack_model((X_1.shape[1],X_1.shape[2]))

X_1 = pickle.load(open("pickled_vectors/mfcc_coefficients_training_vector.pickle","rb"))
X_test_1 = pickle.load(open("pickled_vectors/mfcc_coefficients_evaluation_training_vector.pickle","rb"))

X_2 = pickle.load(open("pickled_vectors/spectral-contrast_peaks_training_vector.pickle","rb"))
X_test_2 = pickle.load(open("pickled_vectors/spectral-contrast_peaks_evaluation_training_vector.pickle","rb"))

model_1 = mfcc_model.mfcc_model((X_1.shape[1],X_1.shape[2]))



model_2 = spectral_contrast_peaks_model.model((X_2.shape[1],X_2.shape[2]))

print("y",y.shape)
print("y_test",y_test.shape)
# print("X_1",X_1.shape)
# print("X_test_1",X_test_1.shape)
print("X_1",X_1.shape)
print("X_test_1",X_test_1.shape)
print("X_2",X_2.shape)
print("X_test_2",X_test_2.shape)
#
#
merged = Merge([model_1,model_2],mode="concat")

final_model = Sequential()
final_model.add(merged)
final_model.add(Dense(1000))
final_model.add(Dense(numGenres, activation='softmax'))
final_model.compile(loss='categorical_crossentropy',
              optimizer='adam',
              metrics=['accuracy']
              )

# plot(model_1,to_file="model_1.png")
# plot(model_1,to_file="model_1.png")
# plot(final_model,to_file="merged_model.png")



json_string = final_model.to_json()
with open("model_architecture/merged_model_architecture.json","w") as f:
    f.write(json.dumps(json_string, sort_keys=True,indent=4, separators=(',', ': ')))

print("Fitting")
# #
# final_model.load_weights("model_weights/embeddings_10_sec_split_gztan_merged_model_weights.hdf5")
# #
# # for i in range(10):
# #     print("epoch",i)

history = final_model.fit([X_1,X_2], y,
                            batch_size=batch_size,
                            nb_epoch=nb_epoch,
                            validation_data=([X_test_1,X_test_2], y_test),
                            shuffle="batch"
                            )
if not os.path.exists("model_weights"):
    os.makedirs("model_weights")
final_model.save_weights("model_weights/merged_model_weights.hdf5",overwrite=True)

with open("experimental_results.json","w") as f:
    f.write(json.dumps(history.history, sort_keys=True,indent=4, separators=(',', ': ')))


for k,v in history.history.items():
    _keys = list(history.history.keys())
    _keys.sort()
    plt.subplot(411+_keys.index(k))
    plt.title(k)

    plt.plot(range(0,len(v)),v,marker="8",linewidth=1.5)

plt.tight_layout(pad=0.4, w_pad=0.5, h_pad=1.0)
plt.show()