Used Deepface from main branch and fixed the following: removed TTS and added Timer

main.py

 #model.h5 is best model
 import cv2
 import warnings
+import pyttsx3
 warnings.filterwarnings("ignore")
-from keras.models import  load_model
-import numpy as np
 import time
+from deepface import DeepFace
+
+# Intiliazing text to speech
+engine = pyttsx3.init('sapi5')
+# Defining camera and starting video capture
+video_capture = cv2.VideoCapture(0)
+# Defining which cascade opencv should use
+face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
+
 
 class Main:
-    def main(self):
-        face_classifier = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')
-        classifier = load_model(r'model.h5')
+    # Welcome Message
+    def welcome_message(self):
+        # Intiliazing text to speech
+
 
-        # Angepasste emotion_labels Liste ohne 'Angry', 'Disgust', 'Fear'
-        emotion_labels = ['Happy', 'Neutral', 'Sad', 'Surprise']
+        engine.say('Welcome to the face recognition and prediction tool of the herman hollerith center')
+        engine.runAndWait()
 
-        cap = cv2.VideoCapture(0)
+    def face_recognition(self):
 
-        start_time = time.time()
-        emotion_history = []
+        # Intiliazing text to speech
+        engine = pyttsx3.init('sapi5')
+        last_analysis_time = time.time()  # Initialisiere die letzte Analysezeit
+        analysis_interval = 4
+        analyze = None
 
         while True:
-            _, frame = cap.read()
+            ret, frame = video_capture.read()
             gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
-            faces = face_classifier.detectMultiScale(gray)
+            faces = face_cascade.detectMultiScale(gray, scaleFactor=1.1, minNeighbors=5)
 
-            for (x, y, w, h) in faces:
-                cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 255), 2)
-                roi_gray = gray[y:y + h, x:x + w]
-                roi_gray = cv2.resize(roi_gray, (48, 48), interpolation=cv2.INTER_AREA)
+            current_time = time.time()
 
-                if np.sum([roi_gray]) != 0:
-                    roi = roi_gray.astype('float') / 255.0
-                    roi = np.expand_dims(roi, axis=0)
+            if current_time - last_analysis_time > analysis_interval:
+                try:
+                    # Annahme: DeepFace.analyze gibt eine Liste zurück. Zugriff auf das erste Element.
+                    analysis_results = DeepFace.analyze(frame, actions=["emotion", "age", "gender"],enforce_detection=False)
+                    analyze = analysis_results[0] if analysis_results else None
+                    last_analysis_time = current_time
+                except Exception as e:
+                    print("Error in DeepFace Analysis:", e)
+                    analyze = None  # Bei Fehlern wird `analyze` zurückgesetzt
 
-                    prediction = classifier.predict(roi)[0]
-                    adjusted_prediction = np.delete(prediction, [0, 1, 2])
-                    max_index = np.argmax(adjusted_prediction)
-                    current_emotion = emotion_labels[max_index]
-                    label_position = (x, y)
-                    cv2.putText(frame, current_emotion, label_position, cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)
+            for (x, y, w, h) in faces:
+                cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 0, 255), 1)
+                if analyze:
+                    # Zugriff auf die Analyseergebnisse, wenn `analyze` nicht None ist
+                    cv2.putText(frame, f'Approx. Age: {analyze.get("age", "N/A")}', (x, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 0, 255), 2)
+                    cv2.putText(frame, f'Approx. Gender: {analyze.get("dominant_gender", "N/A")}', (x, y - 40), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 255), 2)
+                    cv2.putText(frame, f'Current emotion: {analyze.get("dominant_emotion", "N/A")}', (x, y - 70), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (255, 255, 0), 2)
+
+            cv2.imshow("video_capture", frame)
+            if cv2.waitKey(1) & 0xFF == ord('q'):
+                break
 
-                    emotion_history.append(current_emotion)
+        video_capture.release()
+        cv2.destroyAllWindows()
 
-                    # Überprüfen, ob 3 Sekunden vergangen sind diese wird dann für die chat gpt api genutzt
-                    if time.time() - start_time >= 3:
-                        if emotion_history.count(current_emotion) == len(emotion_history):
-                            print(f"Konstante Emotion für 3 Sekunden: {current_emotion}")
 
-                        start_time = time.time()
-                        emotion_history.clear()
+if __name__ == '__main__':
+    main = Main()
 
-            cv2.imshow('Emotion Detector', frame)
-            if cv2.waitKey(1) & 0xFF == ord('q'):
-                break
+    #main.welcome_message()
+    main.face_recognition()
 
-        cap.release()
-        cv2.destroyAllWindows()
 
-if __name__ == "__main__":
-    main_instance = Main()
-    main_instance.main()