Update app.py

app.py

@@ -4,10 +4,10 @@ import streamlit as st
 from PIL import Image
 import numpy as np
 import cv2
+import dlib
 from keras.models import load_model
 from datetime import datetime
 from huggingface_hub import HfApi
-import face_recognition
 
 # Constants
 KNOWN_FACES_DIR = "known_faces"  # Directory to save user images
@@ -30,7 +30,9 @@ REPO_TYPE = "space"  # 'space' type for Streamlit-based projects
 # Load emotion detection model
 model = load_model('CNN_Model_acc_75.h5')
 emotion_labels = ['angry', 'fear', 'happy', 'neutral', 'sad', 'surprise']
-face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')
+face_detector = dlib.get_frontal_face_detector()  # Use Dlib's face detector
+face_predictor = dlib.shape_predictor('shape_predictor_68_face_landmarks.dat')  # Landmarks model for face recognition
+face_rec_model = dlib.face_recognition_model_v1('dlib_face_recognition_resnet_model_v1.dat')  # Face recognition model
 
 # Initialize the SQLite database
 def initialize_database():
@@ -92,19 +94,18 @@ def save_image_to_hugging_face(image, name, roll_no):
 # Process each frame for emotion detection and face recognition
 def process_frame(frame, known_face_encodings, known_face_names):
     gray_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
-    faces = face_cascade.detectMultiScale(gray_frame, scaleFactor=1.1, minNeighbors=5, minSize=(30, 30))
+    faces = face_detector(gray_frame)
 
     face_locations = []
     face_encodings = []
     face_names = []
-    for (x, y, w, h) in faces:
-        roi_color = frame[y:y+h, x:x+w]
-        # Face recognition
-        rgb_frame = frame[:, :, ::-1]  # Convert to RGB (face_recognition expects RGB)
-        face_encoding = face_recognition.face_encodings(rgb_frame, [(y, x + w, y + h, x)])[0]
+    for face in faces:
+        # Get the landmarks of the face
+        shape = face_predictor(gray_frame, face)
+        face_encoding = np.array(face_rec_model.compute_face_descriptor(frame, shape))
 
         # Check if the detected face matches any known face
-        matches = face_recognition.compare_faces(known_face_encodings, face_encoding)
+        matches = [np.allclose(face_encoding, known_encoding) for known_encoding in known_face_encodings]
         name = "Unknown"
 
         if True in matches:
@@ -112,6 +113,7 @@ def process_frame(frame, known_face_encodings, known_face_names):
             name = known_face_names[first_match_index]
 
         # Emotion detection
+        roi_color = frame[face.top():face.bottom(), face.left():face.right()]
         face_roi = cv2.resize(roi_color, (48, 48))
         face_roi = np.expand_dims(face_roi, axis=0)
         face_roi = face_roi / float(48)
@@ -119,8 +121,8 @@ def process_frame(frame, known_face_encodings, known_face_names):
         emotion = emotion_labels[np.argmax(predictions[0])]
 
         # Display name and emotion text on face
-        cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
-        cv2.putText(frame, f"{name} - {emotion}", (x, y+h), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)
+        cv2.rectangle(frame, (face.left(), face.top()), (face.right(), face.bottom()), (0, 255, 0), 2)
+        cv2.putText(frame, f"{name} - {emotion}", (face.left(), face.bottom() + 20), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)
 
     return frame
 
@@ -141,8 +143,8 @@ rows = cursor.fetchall()
 
 for row in rows:
     name, image_path = row
-    image = face_recognition.load_image_file(image_path)
-    encoding = face_recognition.face_encodings(image)[0]
+    image = Image.open(image_path)
+    encoding = dlib.face_recognition_model_v1.compute_face_descriptor(image)[0]
     known_face_encodings.append(encoding)
     known_face_names.append(name)
 
@@ -181,7 +183,7 @@ if st.button("Register"):
             save_to_database(name, roll_no, image_path)
 
             # Update the known faces list
-            known_face_encodings.append(face_recognition.face_encodings(image)[0])
+            known_face_encodings.append(dlib.face_recognition_model_v1.compute_face_descriptor(image)[0])
             known_face_names.append(name)
 
             st.success(f"Student {name} registered successfully!")