Merge pull request #1155 from shivenyadavs/master

Added Face emotion classification Training model

Testemotion.py

@@ -0,0 +1,200 @@
+# import cv2
+# import numpy as np
+# from keras.models import model_from_json
+
+
+# emotion_dict = {0: "Angry", 1: "Disgusted", 2: "Fearful", 3: "Happy", 4: "Neutral", 5: "Sad", 6: "Surprised"}
+
+# # load json and create model
+# json_file = open('emotion_model.json', 'r')
+# loaded_model_json = json_file.read()
+# json_file.close()
+# emotion_model = model_from_json(loaded_model_json)
+
+# # load weights into new model
+# emotion_model.load_weights("emotion_model.h5")
+# print("Loaded model from disk")
+
+# # start the webcam feed
+# #cap = cv2.VideoCapture(0)
+
+# # pass here your video path
+# # you may download one from here : https://www.pexels.com/video/three-girls-laughing-5273028/
+# cap = cv2.VideoCapture(r"C:\Users\Admin\emotiontrain\WIN_20240630_00_58_04_Pro.mp4")
+
+# while True:
+#     # Find haar cascade to draw bounding box around face
+#     ret, frame = cap.read()
+#     frame = cv2.resize(frame, (1280, 720))
+#     if not ret:
+#         break
+#     face_detector = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
+#     gray_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+
+#     # detect faces available on camera
+#     num_faces = face_detector.detectMultiScale(gray_frame, scaleFactor=1.3, minNeighbors=5)
+
+#     # take each face available on the camera and Preprocess it
+#     for (x, y, w, h) in num_faces:
+#         cv2.rectangle(frame, (x, y-50), (x+w, y+h+10), (0, 255, 0), 4)
+#         roi_gray_frame = gray_frame[y:y + h, x:x + w]
+#         cropped_img = np.expand_dims(np.expand_dims(cv2.resize(roi_gray_frame, (48, 48)), -1), 0)
+
+#         # predict the emotions
+#         emotion_prediction = emotion_model.predict(cropped_img)
+#         maxindex = int(np.argmax(emotion_prediction))
+#         cv2.putText(frame, emotion_dict[maxindex], (x+5, y-20), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 0, 0), 2, cv2.LINE_AA)
+
+#     cv2.imshow('Emotion Detection', frame)
+#     if cv2.waitKey(1) & 0xFF == ord('q'):
+#         break
+
+# cap.release()
+# cv2.destroyAllWindows()
+
+
+
+import cv2
+import numpy as np
+from keras.models import model_from_json
+
+emotion_dict = {0: "Angry", 1: "Disgusted", 2: "Fearful", 3: "Happy", 4: "Neutral", 5: "Sad", 6: "Surprised"}
+
+# Load json and create model
+json_file = open('emotion_model.json', 'r')
+loaded_model_json = json_file.read()
+json_file.close()
+emotion_model = model_from_json(loaded_model_json)
+
+# Load weights into new model
+emotion_model.load_weights("final_train.h5")
+print("Loaded model from disk")
+
+# Start the webcam feed
+cap = cv2.VideoCapture(0)
+
+# Pass here your video path
+#cap = cv2.VideoCapture(r"C:\Users\Admin\emotiontrain\WIN_20240630_00_58_04_Pro.mp4")
+
+while True:
+    # Find Haar cascade to draw bounding box around face
+    ret, frame = cap.read()
+    frame = cv2.resize(frame, (1280, 720))
+    if not ret:
+        break
+    face_detector = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
+    gray_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+
+    # Detect faces available on camera
+    num_faces = face_detector.detectMultiScale(gray_frame, scaleFactor=1.3, minNeighbors=5)
+
+    # Take each face available on the camera and preprocess it
+    for (x, y, w, h) in num_faces:
+        # Adjust the rectangle to make it slightly larger
+        padding = 20
+        cv2.rectangle(frame, (x - padding, y - 40 - padding), (x + w + padding, y + h + 10 + padding), (0, 255, 0), 4)
+
+        roi_gray_frame = gray_frame[y:y + h, x:x + w]
+        cropped_img = np.expand_dims(np.expand_dims(cv2.resize(roi_gray_frame, (48, 48)), -1), 0)
+
+        # Predict the emotions
+        emotion_prediction = emotion_model.predict(cropped_img)
+        maxindex = int(np.argmax(emotion_prediction))
+
+        # Adjust the text position to come out of the box
+        cv2.putText(frame, emotion_dict[maxindex], (x + 5, y - 80), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 0, 0), 2, cv2.LINE_AA)
+
+    cv2.imshow('Emotion Detection', frame)
+    if cv2.waitKey(1) & 0xFF == ord('q'):
+        break
+
+cap.release()
+cv2.destroyAllWindows()
+
+# import cv2
+# import numpy as np
+# from keras.models import model_from_json
+
+# # Load the emotion recognition model
+# emotion_dict = {0: "Angry", 1: "Disgusted", 2: "Fearful", 3: "Happy", 4: "Neutral", 5: "Sad", 6: "Surprised"}
+
+# # Load json and create model
+# json_file = open('emotion_model.json', 'r')
+# loaded_model_json = json_file.read()
+# json_file.close()
+# emotion_model = model_from_json(loaded_model_json)
+
+# # Load weights into new model
+# emotion_model.load_weights("emotion_model.h5")
+# print("Loaded model from disk")
+
+# # Start the webcam feed or video file
+# # cap = cv2.VideoCapture(0)  # For webcam
+# cap = cv2.VideoCapture(r"C:\Users\Admin\emotiontrain\WIN_20240630_00_58_04_Pro.mp4")  # For video file
+
+# while True:
+#     # Read frame from video
+#     ret, frame = cap.read()
+#     if not ret:
+#         break
+#     frame = cv2.resize(frame, (640, 360))  # Resize for display
+
+#     # Convert frame to grayscale for face detection
+#     gray_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+#     face_detector = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
+
+#     # Detect faces in the frame
+#     num_faces = face_detector.detectMultiScale(gray_frame, scaleFactor=1.3, minNeighbors=5)
+
+#     # Create a copy of the frame for segmentation display
+#     segmented_frame = frame.copy()
+
+#     # Process each detected face
+#     for (x, y, w, h) in num_faces:
+#         padding = 20
+#         # Draw rectangle on the original frame
+#         cv2.rectangle(frame, (x - padding, y - 40 - padding), (x + w + padding, y + h + 10 + padding), (0, 255, 0), 2)
+
+#         roi_gray_frame = gray_frame[y:y + h, x:x + w]
+#         cropped_img = np.expand_dims(np.expand_dims(cv2.resize(roi_gray_frame, (48, 48)), -1), 0)
+
+#         # Predict emotion
+#         emotion_prediction = emotion_model.predict(cropped_img)
+#         maxindex = int(np.argmax(emotion_prediction))
+#         emotion_label = emotion_dict[maxindex]
+
+#         # Draw emotion label on both frames
+#         cv2.putText(frame, emotion_label, (x + 5, y - 80), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 0, 0), 2, cv2.LINE_AA)
+#         cv2.putText(segmented_frame, emotion_label, (x + 5, y - 80), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 0, 0), 2, cv2.LINE_AA)
+
+#         # Segment the detected face region
+#         face_region = segmented_frame[y:y+h, x:x+w]
+#         lab_face_region = cv2.cvtColor(face_region, cv2.COLOR_BGR2LAB)
+#         pixel_values = lab_face_region.reshape((-1, 3))
+#         pixel_values = np.float32(pixel_values)
+
+#         # K-means clustering
+#         criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 100, 0.2)
+#         k = 3  # Number of clusters
+#         _, labels, (centers) = cv2.kmeans(pixel_values, k, None, criteria, 10, cv2.KMEANS_RANDOM_CENTERS)
+#         centers = np.uint8(centers)
+#         segmented_face_region = centers[labels.flatten()]
+#         segmented_face_region = segmented_face_region.reshape(lab_face_region.shape)
+#         segmented_face_region_bgr = cv2.cvtColor(segmented_face_region, cv2.COLOR_LAB2BGR)
+
+#         # Replace the original face region with the segmented face region
+#         segmented_frame[y:y+h, x:x+w] = segmented_face_region_bgr
+
+#     # Display the results in two separate windows
+#     cv2.imshow('Original Emotion Detection', frame)
+#     cv2.imshow('Segmented Emotion Detection', segmented_frame)
+
+#     # Position the windows next to each other
+#     cv2.moveWindow('Original Emotion Detection', 0, 0)
+#     cv2.moveWindow('Segmented Emotion Detection', 650, 0)  # Adjust the x-coordinate for positioning
+
+#     if cv2.waitKey(1) & 0xFF == ord('q'):
+#         break
+
+# cap.release()
+# cv2.destroyAllWindows()

Training.ipynb

@@ -0,0 +1,69 @@
+
+# import required packages
+import cv2
+from keras.models import Sequential
+from keras.layers import Conv2D, MaxPooling2D, Dense, Dropout, Flatten
+from tensorflow.keras.optimizers import Adam
+from tensorflow.keras.preprocessing.image import ImageDataGenerator
+
+# Initialize image data generator with rescaling
+train_data_gen = ImageDataGenerator(rescale=1./255)
+validation_data_gen = ImageDataGenerator(rescale=1./255)
+
+# Preprocess all test images
+train_generator = train_data_gen.flow_from_directory(
+        'train',
+        target_size=(48, 48),
+        batch_size=64,
+        color_mode="grayscale",
+        class_mode='categorical')
+
+# Preprocess all train images
+validation_generator = validation_data_gen.flow_from_directory(
+        'test',
+        target_size=(48, 48),
+        batch_size=64,
+        color_mode="grayscale",
+        class_mode='categorical')
+
+# create model structure
+emotion_model = Sequential()
+
+emotion_model.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(48, 48, 1)))
+emotion_model.add(Conv2D(64, kernel_size=(3, 3), activation='relu'))
+emotion_model.add(MaxPooling2D(pool_size=(2, 2)))
+emotion_model.add(Dropout(0.25))
+
+emotion_model.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))
+emotion_model.add(MaxPooling2D(pool_size=(2, 2)))
+emotion_model.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))
+emotion_model.add(MaxPooling2D(pool_size=(2, 2)))
+emotion_model.add(Dropout(0.25))
+
+emotion_model.add(Flatten())
+emotion_model.add(Dense(1024, activation='relu'))
+emotion_model.add(Dropout(0.5))
+emotion_model.add(Dense(7, activation='softmax'))
+
+cv2.ocl.setUseOpenCL(False)
+
+emotion_model.compile(loss='categorical_crossentropy', optimizer=Adam(learning_rate=0.0001, decay=1e-6), metrics=['accuracy'])
+
+# Train the neural network/model
+emotion_model_info = emotion_model.fit(
+    train_generator,
+    steps_per_epoch=28709 // 64,
+    epochs=50,
+    validation_data=validation_generator,
+    validation_steps=7178 // 64
+)
+
+
+# save model structure in jason file
+model_json = emotion_model.to_json()
+with open("emotion_model.json", "w") as json_file:
+    json_file.write(model_json)
+
+# save trained model weight in .h5 file
+emotion_model.save('emotion_model.h5')
+

Training.py

@@ -0,0 +1,69 @@
+
+# import required packages
+import cv2
+from keras.models import Sequential
+from keras.layers import Conv2D, MaxPooling2D, Dense, Dropout, Flatten
+from tensorflow.keras.optimizers import Adam
+from tensorflow.keras.preprocessing.image import ImageDataGenerator
+
+# Initialize image data generator with rescaling
+train_data_gen = ImageDataGenerator(rescale=1./255)
+validation_data_gen = ImageDataGenerator(rescale=1./255)
+
+# Preprocess all test images
+train_generator = train_data_gen.flow_from_directory(
+        'train',
+        target_size=(48, 48),
+        batch_size=64,
+        color_mode="grayscale",
+        class_mode='categorical')
+
+# Preprocess all train images
+validation_generator = validation_data_gen.flow_from_directory(
+        'test',
+        target_size=(48, 48),
+        batch_size=64,
+        color_mode="grayscale",
+        class_mode='categorical')
+
+# create model structure
+emotion_model = Sequential()
+
+emotion_model.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(48, 48, 1)))
+emotion_model.add(Conv2D(64, kernel_size=(3, 3), activation='relu'))
+emotion_model.add(MaxPooling2D(pool_size=(2, 2)))
+emotion_model.add(Dropout(0.25))
+
+emotion_model.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))
+emotion_model.add(MaxPooling2D(pool_size=(2, 2)))
+emotion_model.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))
+emotion_model.add(MaxPooling2D(pool_size=(2, 2)))
+emotion_model.add(Dropout(0.25))
+
+emotion_model.add(Flatten())
+emotion_model.add(Dense(1024, activation='relu'))
+emotion_model.add(Dropout(0.5))
+emotion_model.add(Dense(7, activation='softmax'))
+
+cv2.ocl.setUseOpenCL(False)
+
+emotion_model.compile(loss='categorical_crossentropy', optimizer=Adam(learning_rate=0.0001, decay=1e-6), metrics=['accuracy'])
+
+# Train the neural network/model
+emotion_model_info = emotion_model.fit(
+    train_generator,
+    steps_per_epoch=28709 // 64,
+    epochs=50,
+    validation_data=validation_generator,
+    validation_steps=7178 // 64
+)
+
+
+# save model structure in jason file
+model_json = emotion_model.to_json()
+with open("emotion_model.json", "w") as json_file:
+    json_file.write(model_json)
+
+# save trained model weight in .h5 file
+emotion_model.save('emotion_model.h5')
+