bug fixed images - path

.idea/ascii_demo.iml

@@ -1,7 +1,9 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <module type="PYTHON_MODULE" version="4">
   <component name="NewModuleRootManager">
-    <content url="file://$MODULE_DIR$" />
+    <content url="file://$MODULE_DIR$">
+      <sourceFolder url="file://$MODULE_DIR$/pages" isTestSource="false" />
+    </content>
     <orderEntry type="inheritedJdk" />
     <orderEntry type="sourceFolder" forTests="false" />
   </component>

pages/3_Training the Model.py

@@ -3,6 +3,7 @@ from PIL import Image
 from utils.levels import complete_level, render_page, initialize_level
 from utils.login import get_login, initialize_login
 from utils.inference import query
+from utils.database import get_database
 import os
 import time
 import face_recognition
@@ -18,15 +19,6 @@ LEVEL = 3
 
 PKL_PATH = 'dataset/database.pkl'
 
-def get_database():
-    database = {}
-
-    if os.path.getsize(PKL_PATH) > 0:
-        with open(PKL_PATH, "rb") as f:
-            unpickler = pickle.Unpickler(f)
-            database = unpickler.load()
-    return database
-
 def step3_page():
     st.header("Training the Model")
     st.subheader("Face encoding")
@@ -71,7 +63,7 @@ def step3_page():
     if st.button("Train Model"):
         my_bar = st.progress(0, text="Training....")
         if len(images) > 0:
-            database = get_database()
+            database = get_database(PKL_PATH)
             for i in range(100):
                 my_bar.progress(i, text="Training....")
             my_bar.progress(100, text="Successfully Trained!")
@@ -85,23 +77,25 @@ def step3_page():
                 face_id = img.split(".")[0]
 
                 # check if id already exists
-                existing_id = [i for i in database.keys()]
+                existing_id = [database[i]["face_id"] for i in database.keys()]
                 if face_id in existing_id:
                     st.error(f"Encoding already created for : {face_id}")
                 else:
-                    database[face_id] = {'name': face_name,
-                                         'encoding': my_face_encoding}
+                    database[i] = {'face_id': face_id,
+                                   'name': face_name,
+                                   'encoding': my_face_encoding}
                     with open(PKL_PATH, 'wb') as f:
                         pkl.dump(database, f)
 
                 my_bar.progress(int((i + 1) / len(images) * 100), text="Generating face encodings...")
             my_bar.progress(100, text="Successfully encoded all the known faces!")
             st.success("Face encoding completed successfully!")
-            if st.button("Complete"):
-                complete_level(LEVEL)
+
         else:
             my_bar.empty()
             st.error("You have not taken any images yet! Do the previous steps first!")
+    if st.button("Complete"):
+        complete_level(LEVEL)
 
 
 render_page(step3_page, LEVEL)

pages/4_Trying It Out.py

@@ -1,7 +1,8 @@
 import streamlit as st
 from utils.levels import complete_level, render_page, initialize_level
 from utils.login import get_login, initialize_login
-from utils.inference import query
+from utils.inference import recognize
+from utils.database import get_database
 import os
 import time
 import face_recognition
@@ -28,56 +29,43 @@ def step4_page():
     st.info(
         "Now that we know how our face recognition application works, let's try it out!"
     )
-    face_encodings_dir = os.path.join(".sessions", get_login()["username"], "face_encodings")
-    face_encodings = os.listdir(face_encodings_dir)
-    known_face_encodings = []
-    known_face_names = []
-    if len(face_encodings) > 0:
-        for i, face_encoding in enumerate(face_encodings):
-            known_face_encoding = np.load(os.path.join(face_encodings_dir, face_encoding))
-            face_name = face_encoding.split(".")[0]
-            known_face_encodings.append(known_face_encoding)
-            known_face_names.append(face_name)
-
-    st.info("Select an image to analyze!")
+    # Select input type
+    st.info("Select your input type to analyze!")
     input_type = st.radio("Select the Input Type", ["Image", "Camera"])
+    # Put slide to adjust tolerance
+    tolerance = st.slider("Tolerance", 0.0, 1.0, 0.5, 0.01)
+    st.info(
+        "Tolerance is the threshold for face recognition. The lower the tolerance, the more strict the face recognition. The higher the tolerance, the more loose the face recognition.")
 
-    if input_type == "Camera":
-        picture = st.camera_input("Take a picture")
+    if input_type == "Image":
+        st.title("Face Recognition App")
+        uploaded_images = st.file_uploader("Upload", type=['jpg', 'png', 'jpeg'], accept_multiple_files=True)
+        if len(uploaded_images) != 0:
+            # Read uploaded image with face_recognition
+            for image in uploaded_images:
+                image = face_recognition.load_image_file(image)
+                image, name, face_id = recognize(image, tolerance)
+                st.image(image)
+        else:
+            st.info("Please upload an image")
     else:
-        picture = st.file_uploader("Upload an image", type=["png", "jpg", "jpeg"])
-    if picture:
-        image = face_recognition.load_image_file(picture)
-        face_locations = face_recognition.face_locations(image)
-        face_encodings = face_recognition.face_encodings(image, face_locations)
-
-        st.image(image)
-        # Loop through each face in this image
-        cols = st.columns(len(face_encodings))
-        i = 0
-        # st.info("Select the tolerance level you want for your model! (How much distance between faces to consider it a match. "
-        #         "Lower is more strict. 0.6 is typical best performance.)")
-        # tolerance = st.slider('Select tolerance level', 0.0, 1.0, 0.3, 0.1)
-        # if tolerance:
-        for (top, right, bottom, left), face_encoding in zip(face_locations, face_encodings):
-            # See if the face is a match for the known face(s)
-            # matches = face_recognition.compare_faces(known_face_encodings, face_encoding)
-
-            name = "Unknown"
-            # If a match was found in known_face_encodings, just use the first one.
-            face_distances = face_recognition.face_distance(known_face_encodings, face_encoding)
-
-            # Calculate the row sums
-            row_sums = np.sum(face_distances, axis=1)
-            best_match_index = np.argmin(row_sums)
-            if best_match_index is not None:
-                name = known_face_names[best_match_index]
+        st.title("Face Recognition App")
+        # Camera Settings
+        cam = cv2.VideoCapture(0)
+        cam.set(cv2.CAP_PROP_FRAME_WIDTH, 640)
+        cam.set(cv2.CAP_PROP_FRAME_HEIGHT, 480)
+        FRAME_WINDOW = st.image([])
 
-            face_image = image[top:bottom, left:right]
-            pil_image = Image.fromarray(face_image)
-            cols[i].image(pil_image, use_column_width=True)
-            cols[i].write("Person name: " +name)
-            i+=1
+        while True:
+            ret, frame = cam.read()
+            if not ret:
+                st.error("Failed to capture frame from camera")
+                st.info("Please turn off the other app that is using the camera and restart app")
+                st.stop()
+            image, name, face_id = recognize(frame, tolerance)
+            image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
+            # Display name and ID of the person
+            FRAME_WINDOW.image(image)
 
     st.info("Click on the button below to complete this level!")
     if st.button("Complete Level"):

utils/database.py

@@ -0,0 +1,11 @@
+import os
+import pickle as pkl
+
+def get_database(PKL_PATH):
+    database = {}
+
+    if os.path.getsize(PKL_PATH) > 0:
+        with open(PKL_PATH, "rb") as f:
+            unpickler = pickle.Unpickler(f)
+            database = unpickler.load()
+    return database

utils/inference.py

@@ -1,10 +1,28 @@
-import requests
+from utils.database import get_database
+import os
+import face_recognition
+import cv2
+import numpy as np
+from PIL import Image
 
-API_URL = "https://api-inference.huggingface.co/models/CynthiaCR/emotions_classifier"
-headers = {"Authorization": "Bearer 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"}
-
-def query(filename):
-    with open(filename, "rb") as f:
-        data = f.read()
-    response = requests.post(API_URL, headers=headers, data=data)
-    return response.json()
+def recognize(image,tolerance):
+    database = get_databse()
+    known_encoding = [database[id]['encoding'] for id in database.keys()]
+    name = 'Unknown'
+    face_id = 'Unknown'
+    face_locations = face_recognition.face_locations(image)
+    face_encodings = face_recognition.face_encodings(image,face_locations)
+    for (top,right,bottom,left), face_encoding in zip(face_locations,face_encodings):
+        matches = face_recognition.compare_faces(known_encoding,face_encoding,tolerance=tolerance)
+        distance = face_recognition.face_distance(known_encoding,face_encoding)
+        name = 'Unknown'
+        face_id = 'Unknown'
+        if True in matches:
+            match_index = matches.index(True)
+            name = database[match_index]['name']
+            face_id = database[match_index]['face_id'].split("_")[1]
+            distance = round(distance[match_index],2)
+            cv2.putText(image,str(distance),(left,top-30),cv2.FONT_HERSHEY_SIMPLEX,0.75,(0,255,0),2)
+        cv2.rectangle(image,(left,top),(right,bottom),(0,255,0),2)
+        cv2.putText(image,name,(left,top-10),cv2.FONT_HERSHEY_SIMPLEX,0.75,(0,255,0),2)
+    return image, name, face_id