Final Version of the Project

FER/detectfaces.py

@@ -1,39 +1,14 @@
-from models.PosterV2_7cls import pyramid_trans_expr2
 import cv2
 import torch
 import os
 import time
 from PIL import Image
-from main import RecorderMeter1, RecorderMeter  # noqa: F401
 
-script_dir = os.path.dirname(os.path.abspath(__file__))
-
-# Construct the full path to the model file
-model_path = os.path.join(script_dir,"models","checkpoints","raf-db-model_best.pth")
-
-# Determine the available device for model execution
-if torch.backends.mps.is_available():
-    device = "mps"
-elif torch.cuda.is_available():
-    device = "cuda"
-else:
-    device = "cpu"
-
-# Initialize the model with specified image size and number of classes
-model = pyramid_trans_expr2(img_size=224, num_classes=7)
-
-# Wrap the model with DataParallel for potential multi-GPU usage
-model = torch.nn.DataParallel(model)
-
-# Move the model to the chosen device
-model = model.to(device)
-
-# Print the current time
 currtime = time.strftime("%H:%M:%S")
 print(currtime)
 
 
-def main():
+def fer(model_path, device, model):
     # Load the model checkpoint if it exists
     if model_path is not None:
         if os.path.isfile(model_path):
@@ -121,8 +96,3 @@ def imagecapture(model):
     # Release webcam resources and close OpenCV windows
     cap.release()
     cv2.destroyAllWindows()
-
-
-# Execute the main function if the script is run directly
-if __name__ == "__main__":
-    main()

FER/sever.py

@@ -0,0 +1,94 @@
+import socket
+import os 
+import torch 
+from threading import Timer
+import pyttsx3
+import speech_recognition as sr
+from detectfaces import fer
+from models.PosterV2_7cls import pyramid_trans_expr2
+from main import RecorderMeter1, RecorderMeter  # noqa: F401 
+import time
+
+script_dir = os.path.dirname(os.path.abspath(__file__))
+
+# Construct the full path to the model file
+model_path = os.path.join(script_dir,"models","checkpoints","raf-db-model_best.pth")
+
+# Determine the available device for model execution
+if torch.backends.mps.is_available():
+    device = "mps"
+elif torch.cuda.is_available():
+    device = "cuda"
+else:
+    device = "cpu"
+
+# Initialize the model with specified image size and number of classes
+model = pyramid_trans_expr2(img_size=224, num_classes=7)
+
+# Wrap the model with DataParallel for potential multi-GPU usage
+model = torch.nn.DataParallel(model)
+
+# Move the model to the chosen device
+model = model.to(device)
+fer(model_path=model_path, device=device, model=model)
+
+s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+s.bind(('', 5001))
+s.listen(5)
+print("Bot is Running")
+
+def handle_client(clientsocket):
+    r = sr.Recognizer()
+    m = sr.Microphone()
+
+    try:
+        while True:
+            prompt = ''
+            print("Speak now:")
+            sent = False
+            with m as source:
+                audio = r.listen(source)
+
+            try:
+                prompt = r.recognize_google(audio)
+                print("Tadbot Thinks you said:", prompt)
+                sent = True
+            except sr.UnknownValueError:
+                print("Tadbot could not understand audio. Try Again")
+            except sr.RequestError as e:
+                print(f"Could not request results from Google Speech Recognition service: {e}")
+
+            if sent:
+                print("please Wait!")
+                try:
+                    clientsocket.send(bytes(prompt, 'utf-8'))
+                    response = clientsocket.recv(1024).decode("utf-8")
+                    engine = pyttsx3.init('espeak')
+                    voices = engine.getProperty('voices')
+                    engine.setProperty('voice', voices[26].id)
+                    engine.setProperty('rate', 145)
+                    engine.say(response)
+                    engine.runAndWait()
+                    print("TADBot:", response)
+                except (socket.error, ConnectionResetError) as e:
+                    print(f"Connection error: {e}")
+                    break  # Exit loop if connection breaks
+                time.sleep(60)  # Wait for 60 seconds before listening again
+
+    finally:
+        clientsocket.close()
+        print("Connection Closed")
+
+
+while True:
+    try:
+        clientsocket, address = s.accept()
+        print(f"Accepted connection from {address}")
+        handle_client(clientsocket) #Handle each client in a separate function
+    except KeyboardInterrupt:
+        print("Server shutting down...")
+        break
+    except Exception as e:
+        print(f"An error occurred: {e}")
+
+s.close()

FER/test.py

@@ -0,0 +1,31 @@
+from detectfaces import fer
+from models.PosterV2_7cls import pyramid_trans_expr2
+import os 
+import torch 
+from main import RecorderMeter1, RecorderMeter  # noqa: F401 
+script_dir = os.path.dirname(os.path.abspath(__file__))
+
+# Construct the full path to the model file
+model_path = os.path.join(script_dir,"models","checkpoints","raf-db-model_best.pth")
+
+# Determine the available device for model execution
+if torch.backends.mps.is_available():
+    device = "mps"
+elif torch.cuda.is_available():
+    device = "cuda"
+else:
+    device = "cpu"
+
+# Initialize the model with specified image size and number of classes
+model = pyramid_trans_expr2(img_size=224, num_classes=7)
+
+# Wrap the model with DataParallel for potential multi-GPU usage
+model = torch.nn.DataParallel(model)
+
+# Move the model to the chosen device
+model = model.to(device)
+def main():
+    fer(model_path=model_path, device=device, model=model)
+
+if __name__ == "__main__":
+    main()

Output.py

@@ -1,3 +1,4 @@
+
 import pyttsx3
 
 engine = pyttsx3.init('espeak')
@@ -5,6 +6,7 @@ voices = engine.getProperty('voices')
 print("Available voices:")
 for voice in voices:
     print(f"- {voice.id}")
-engine.setProperty('voice', voices[14].id)  # Use the first available voice
-engine.say("Hello from Raspberry Pi")
+engine.setProperty('voice', voices[29].id)
+engine.setProperty('rate', 145)  # Use the first available voice
+engine.say("Hi, I AM A DONKEY")
 engine.runAndWait()

app.py

@@ -0,0 +1,16 @@
+from FER import detectfaces
+from sever import botFunction
+import socket
+from threading import Timer
+# Settin up connection to the server
+s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+s.bind(('',5001))
+s.listen(5)
+
+# Checking if connection exists then perform stuff
+while True: 
+    clientsocket, address = s.accept()
+    detectfaces()
+    botFunction()
+    print("Connection Closed")
+    break