Train.py

import os
import matplotlib.pyplot as plt
from tensorflow.keras.preprocessing.image import ImageDataGenerator
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Dropout
from tensorflow.keras.optimizers import Adam

# === Settings ===
IMG_SIZE = 48
BATCH_SIZE = 64
EPOCHS = 100
DATA_PATH = 'Emotion_dataset'  # Replace with your actual folder path

# === Load Dataset ===
datagen = ImageDataGenerator(rescale=1./255, validation_split=0.2, horizontal_flip=True)

train_generator = datagen.flow_from_directory(
    DATA_PATH,
    target_size=(IMG_SIZE, IMG_SIZE),
    color_mode='grayscale',
    batch_size=BATCH_SIZE,
    class_mode='categorical',
    subset='training'
)

val_generator = datagen.flow_from_directory(
    DATA_PATH,
    target_size=(IMG_SIZE, IMG_SIZE),
    color_mode='grayscale',
    batch_size=BATCH_SIZE,
    class_mode='categorical',
    subset='validation'
)

# === Build CNN Model ===
model = Sequential([
    Conv2D(32, (3, 3), activation='relu', input_shape=(IMG_SIZE, IMG_SIZE, 1)),
    MaxPooling2D(2, 2),

    Conv2D(64, (3, 3), activation='relu'),
    MaxPooling2D(2, 2),

    Conv2D(128, (3, 3), activation='relu'),
    MaxPooling2D(2, 2),

    Flatten(),
    Dropout(0.5),
    Dense(128, activation='relu'),
    Dense(train_generator.num_classes, activation='softmax')
])

# === Compile ===
model.compile(optimizer=Adam(learning_rate=0.0001), loss='categorical_crossentropy', metrics=['accuracy'])

# === Train ===
history = model.fit(train_generator, validation_data=val_generator, epochs=EPOCHS)

# === Save the Model ===
model.save("emotion.h5")
print("✅ Model saved as 'emotion.h5'")

# === Plot Accuracy ===
plt.plot(history.history['accuracy'], label='Train Accuracy')
plt.plot(history.history['val_accuracy'], label='Validation Accuracy')
plt.legend()
plt.title('Model Accuracy')
plt.show()