Modified CNN_Model.py

IT20201364/CNN_Model.py

+
+import os
+import warnings
+import librosa.display
+import numpy as np
+from keras.layers import Dense, Conv2D, MaxPooling2D, Flatten, Dropout, BatchNormalization
+from keras.models import Sequential
+from keras.utils import pad_sequences
+from keras.utils import to_categorical
+from sklearn.model_selection import train_test_split
+from tqdm import tqdm
+
+warnings.filterwarnings("ignore")
+
+dataList = os.listdir('/RawData')
+classLabels = ('Angry', 'Fear', 'Disgust', 'Happy', 'Sad', 'Surprised', 'Neutral')
+data = []
+labels = []
+
+for number, path in enumerate(tqdm(dataList)):
+    X, sample_rate = librosa.load('C:/Users/dell/Desktop/F_Project/Virtual_Assistant/RawData/' + path, res_type='kaiser_best',
+                                  duration=2.5, sr=22050 * 2, offset=0.5)
+
+    sample_rate = np.array(sample_rate)
+    mfccs = librosa.feature.mfcc(y=X, sr=sample_rate, n_mfcc=39)
+    feature = mfccs
+    data.append(feature)
+
+    if path[6:8] == '01' or path[0:1] == 'n':
+        labels.append(6)
+    if path[6:8] == '02':
+        labels.append(6)
+    if path[6:8] == '03' or path[0:1] == 'h':
+        labels.append(3)
+    if path[6:8] == '04' or path[0:2] == 'sa':
+        labels.append(4)
+    if path[6:8] == '05' or path[0:1] == 'a':
+        labels.append(0)
+    if path[6:8] == '06' or path[0:1] == 'f':
+        labels.append(1)
+    if path[6:8] == '07' or path[0:1] == 'd':
+        labels.append(2)
+    if path[6:8] == '08' or path[0:2] == 'su':
+        labels.append(5)
+
+max_len = 216
+data = np.array([pad_sequences(x, maxlen=max_len, padding='post', truncating='post') for x in data])
+
+
+labels = np.array(labels)
+
+X_train, X_test, Y_train, Y_test = train_test_split(data, labels, test_size=0.3, random_state=42)
+
+numLabels = len(classLabels)
+
+Y_train = to_categorical(Y_train)
+Y_test = to_categorical(Y_test)
+
+X_train = X_train[..., np.newaxis]
+X_test = X_test[..., np.newaxis]
+
+model = Sequential()
+model.add(Conv2D(32, (3, 3), activation='relu', input_shape=(X_train.shape[1:])))
+model.add(BatchNormalization())
+model.add(MaxPooling2D(pool_size=(2, 2)))
+
+model.add(Conv2D(64, (3, 3), activation='relu'))
+model.add(BatchNormalization())
+model.add(MaxPooling2D(pool_size=(2, 2)))
+
+model.add(Conv2D(128, (3, 3), activation='relu'))
+model.add(BatchNormalization())
+model.add(MaxPooling2D(pool_size=(2, 2)))
+
+model.add(Flatten())
+model.add(Dense(256, activation='relu'))
+model.add(Dropout(0.3))
+model.add(Dense(numLabels, activation='softmax'))
+
+model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
+print(model.summary())
+
+best_acc = 0
+epochs = 50
+
+for i in tqdm(range(epochs)):
+    model.fit(X_train, Y_train, batch_size=32, epochs=1)
+    loss, acc = model.evaluate(X_test, Y_test)
+    if acc > best_acc:
+        best_acc = acc
+
+model.evaluate(X_test, Y_test)
+print("Best Accuracy:", best_acc)
+model.save("my_model1.h5")