I have created a project and implemented the following basic functions:

- Text to speech
- Video capture
- Facial recognition and analysis
- Age, gender and emotion recognition

I also added a requirements txt that can be used to acquire all used packages via this comment: pip install -r requirements.txt

Emotion Recognition.py

+import cv2
+from deepface import DeepFace
+import pyttsx3
+import numpy as np
+#Global Variables
+
+# 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')
+
+#Welcome Message
+def welcome_message():
+    engine.say('Welcome to the face recognition and prediction tool of the herman hollerith center')
+    engine.runAndWait()
+
+# Face_recognition and prediction function (maybe we should split that up - Single Responsibility principle)
+def face_recognition():
+        while video_capture.isOpened():
+            ret, frame = video_capture.read()
+
+            gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+            faces = face_cascade.detectMultiScale(gray, scaleFactor=1.1, minNeighbors=5)
+            analyze = DeepFace.analyze(frame, actions=["emotion", "age", "gender"])
+            for (x, y, w, h) in faces:
+                img = cv2.rectangle(frame, (x, y), (x+w, y+h), (0, 0, 255), 1)
+                try:
+                    #Adding the results of the prediction into the captured video
+                    cv2.putText(frame,
+                                    f'Approx. Age: {analyze[0]["age"]}',
+                                    (x, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 0, 255), 2)
+                    cv2.putText(frame,
+                                    f'Approx. Gender:{analyze[0]["dominant_gender"]}',
+                                    (x, y - 40), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 255), 2)
+                    cv2.putText(frame,
+                                    f'Current emotion:{analyze[0]["dominant_emotion"]}',
+                                    (x, y - 60), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (255, 255, 0), 2)
+                    # Text to speech the results
+                    engine.say(f'We have predicted that your age is {analyze[0]["age"]}, your gender is {analyze[0]["dominant_gender"]} and your current emotional state is {analyze[0]["dominant_emotion"]}' )
+                    engine.runAndWait()
+                except:
+                    print("No face detected")
+
+            #Showing the video
+            cv2.imshow("video_capture", frame)
+            # Adding a backout key and the time he should wait while showing the current picture
+            if cv2.waitKey(1) & 0xFF == ord('q'):
+                break
+        video_capture.release()
+
+if __name__ == '__main__':
+    welcome_message()
+    face_recognition()
\ No newline at end of file