FIX: Added explicit secondary STUN server to resolve NoneType transport error in aioice.

app.py

@@ -1,60 +1,83 @@
 import streamlit as st
-import cv2  # Essential for processing video frames
+import cv2  # OpenCV for image processing
 import numpy as np
-from streamlit_webrtc import webrtc_streamer, VideoTransformerBase, WebRtcMode
+import time
+from streamlit_webrtc import webrtc_streamer, VideoTransformerBase, WebRtcMode, VideoProcessorBase
+# NOTE: Make sure these core libraries are in your requirements.txt
+# import deepface 
+# import sklearn # if needed for recognition/clustering
 
-# --- [PLACEHOLDER IMPORTS FOR YOUR PROJECT STRUCTURE] ---
-# NOTE: Replace these with your actual imports if using src/ files
+# --- PLACEHOLDER IMPORTS (UNCOMMENT/ADJUST AS NEEDED) ---
 # from src.detect import detect_faces
+# from src.embed import get_embeddings
 # from src.recognize import recognize_face
-# from src.log import LogManager
+# from src.utils import LogManager 
 
-# --- CONFIGURATION (Move to a separate config file if complex) ---
-# Assuming you have a list of known users for recognition
-KNOWN_USERS = ["Adharsh", "Jane Doe", "Guest"] 
-FRAME_SKIP = 5 # Process every 5th frame for performance
 
-# --- VIDEO PROCESSING CLASS ---
+# --- CONFIGURATION ---
+# NOTE: Adjust these values based on your model/system performance
+RECOGNITION_THRESHOLD = 0.6
+FRAME_SKIP = 3  # Process every 3rd frame for performance
+
 
+# --- VIDEO PROCESSING CLASS ---
 # VideoTransformerBase handles receiving frames and sending them back
 class FaceRecognitionTransformer(VideoTransformerBase):
     """
     A class that processes video frames in real-time for face recognition.
     """
-    def __init__(self, recognition_threshold=0.6):
-        # Initialize any models or trackers here
-        # Example: self.model = load_recognition_model() 
-        self.recognition_threshold = recognition_threshold
-        # Placeholder for face detection/recognition models
-        self.detector = None 
-        self.recognizer = None
-
+    def __init__(self):
+        # Initialize any models/trackers here to load them once
+        # Example: self.detector = deepface.DeepFace.build_model("mtcnn")
+        # Example: self.recognizer = load_your_recognizer_model()
+        self.frame_count = 0
+        self.detection_model = None # Placeholder
+        self.recognition_model = None # Placeholder
+        
+        # Log manager placeholder
+        # self.log_manager = LogManager() 
+        
     def transform(self, frame: np.ndarray) -> np.ndarray:
-        # Convert the frame from BGR (OpenCV default) to RGB 
+        # Increment frame count
+        self.frame_count += 1
+        
+        # Skip frames to reduce CPU load
+        if self.frame_count % FRAME_SKIP != 0:
+            return frame
+        
+        # Convert frame from BGR (OpenCV default) to RGB 
         img = frame.copy()
         
-        # 1. Detect Faces
-        # This function should return bounding boxes and confidence scores
-        # Example: faces = detect_faces(img)
-        faces = [] # Placeholder for detected faces (x, y, w, h)
+        # 1. Detect Faces (Placeholder Logic)
+        # Replace with your actual detection function call
+        # Example: faces = detect_faces(img, self.detection_model)
+        
+        # Placeholder: Assume one face in the middle for demonstration
+        # In a real app, you'd get (x, y, w, h) for all faces
+        h, w, _ = img.shape
+        faces = [(w//4, h//4, w//2, h//2)] 
 
         for (x, y, w, h) in faces:
-            # 2. Recognize Face
-            # recognized_name = recognize_face(img, x, y, w, h, self.recognizer)
+            # 2. Recognize Face (Placeholder Logic)
+            # Example: recognized_name, score = recognize_face(img, x, y, w, h, self.recognition_model)
             recognized_name = "Unknown" # Placeholder result
-
-            # 3. Log/Display Result
-            if recognized_name != "Unknown":
+            score = 0.0
+            
+            # --- Decision and Visualization ---
+            
+            if recognized_name != "Unknown" and score >= RECOGNITION_THRESHOLD:
                 color = (0, 255, 0) # Green for known user
-                # LogManager.log_access(recognized_name)
+                # self.log_manager.log_access(recognized_name)
             else:
                 color = (0, 0, 255) # Red for unknown user
+                recognized_name = "Unknown" 
             
             # Draw bounding box
             cv2.rectangle(img, (x, y), (x + w, y + h), color, 2)
             
             # Draw label
-            cv2.putText(img, recognized_name, (x, y - 10), 
+            label = f"{recognized_name}: {score:.2f}"
+            cv2.putText(img, label, (x, y - 10), 
                         cv2.FONT_HERSHEY_SIMPLEX, 0.9, color, 2)
 
         return img
@@ -62,23 +85,32 @@ class FaceRecognitionTransformer(VideoTransformerBase):
 # --- STREAMLIT UI ---
 
 def main():
+    global RECOGNITION_THRESHOLD
+    
     st.title("Smart Office Face Recognition System 📸")
     st.sidebar.title("Configuration")
 
-    # Sidebar controls (optional)
-    recognition_threshold = st.sidebar.slider(
+    # Sidebar control for threshold
+    RECOGNITION_THRESHOLD = st.sidebar.slider(
         "Recognition Threshold", min_value=0.0, max_value=1.0, value=0.6, step=0.05
     )
 
     # Start the WebRTC Streamer
-    # NOTE: Set WebRtcMode.SENDONLY if you only need the camera feed
     webrtc_streamer(
         key="face-recognition-stream",
+        # Use SENDRECV mode for two-way communication (video in, video out)
         mode=WebRtcMode.SENDRECV,
+        
+        # --- CRITICAL FIX: Enhanced STUN/TURN configuration to fix aioice errors ---
         rtc_configuration={
-            "iceServers": [{"urls": ["stun:stun.l.google.com:19302"]}]
+            "iceServers": [
+                # Google's public STUN server (standard)
+                {"urls": ["stun:stun.l.google.com:19302"]},
+                # Mozilla's public STUN server (as backup)
+                {"urls": ["stun:stun.services.mozilla.com"]}
+            ]
         },
-        video_transformer_factory=lambda: FaceRecognitionTransformer(recognition_threshold),
+        video_transformer_factory=FaceRecognitionTransformer,
         async_transform=True
     )
 
@@ -86,7 +118,7 @@ def main():
     st.subheader("Access Log (Placeholder)")
     # Placeholder for displaying logs
     # if st.button("Refresh Log"):
-    #    st.dataframe(LogManager.get_logs())
+    #     st.dataframe(LogManager().get_logs())
 
 # --- EXECUTION ---
 if __name__ == "__main__":