Upload 2 files

app.py

@@ -1,6 +1,6 @@
 from mtcnn.mtcnn import MTCNN
 import streamlit as st
-import matplotlib.pyplot as plt 
+import matplotlib.pyplot as plt
 from matplotlib.patches import Rectangle
 from matplotlib.patches import Circle
 from PIL import Image
@@ -11,18 +11,25 @@ from keras_vggface.vggface import VGGFace
 
 st.header("Face Detection using a Pre-trained CNN model")
 
-choice = st.selectbox("",[
-    "Face Detection - Show Bounding Box",
-    "Face Detection - Extract Face",
-    "Face Verification"
-])
+choice = st.selectbox(
+    "Choose task",
+    [
+        "Face Detection - Show Bounding Box",
+        "Face Detection - Extract Face",
+        "Face Verification",
+    ],
+)
+
+
 def main():
     fig = plt.figure()
     if choice == "Face Detection - Show Bounding Box":
         st.subheader("Face Detection - Show Bounding Box")
-        st.write("Please upload an image containing a face. A box will be drawn highlighting the face using the pretrained VGGFace model.")
+        st.write(
+            "Please upload an image containing a face. A box will be drawn highlighting the face using the pretrained VGGFace model."
+        )
         # load the image
-        uploaded_file = st.file_uploader("Upload Image", type=["jpg","png"], key="1")
+        uploaded_file = st.file_uploader("Upload Image", type=["jpg", "png"], key="1")
         if uploaded_file is not None:
             data = asarray(Image.open(uploaded_file))
             # plot the image
@@ -38,25 +45,29 @@ def main():
             faces = detector.detect_faces(data)
             for face in faces:
                 # get coordinates
-                x, y, width, height = face['box']
+                x, y, width, height = face["box"]
                 # create the shape
-                rect = Rectangle((x, y), width, height, fill=False, color='green')
+                rect = Rectangle((x, y), width, height, fill=False, color="green")
                 # draw the box
                 ax.add_patch(rect)
                 # draw the dots
-                for _, value in face['keypoints'].items():
+                for _, value in face["keypoints"].items():
                     # create and draw dot
-                    dot = Circle(value, radius=2, color='green')
+                    dot = Circle(value, radius=2, color="green")
                     ax.add_patch(dot)
             # show the plot
             st.pyplot(fig)
-        st.write("The box highlights the face and the dots highlight the identified features.")
+        st.write(
+            "The box highlights the face and the dots highlight the identified features."
+        )
 
     elif choice == "Face Detection - Extract Face":
         st.subheader("Face Detection - Extract Face")
-        st.write("Please upload an image containing a face. The part of the image containing the face will be extracted using the pretrained VGGFace model.")
+        st.write(
+            "Please upload an image containing a face. The part of the image containing the face will be extracted using the pretrained VGGFace model."
+        )
 
-        uploaded_file = st.file_uploader("Upload Image", type=["jpg","png"], key="2")
+        uploaded_file = st.file_uploader("Upload Image", type=["jpg", "png"], key="2")
         if uploaded_file is not None:
             column1, column2 = st.columns(2)
             image = Image.open(uploaded_file)
@@ -79,52 +90,78 @@ def main():
             face = pixels[y1:y2, x1:x2]
             # resize pixels to the model size
             image = Image.fromarray(face)
-            image = image.resize((224, 224)) # Rodgers -> You can just save this as image
+            image = image.resize(
+                (224, 224)
+            )  # Rodgers -> You can just save this as image
             face_array = asarray(image)
             with column2:
-                 plt.imshow(face_array)
-                 st.pyplot(fig)
-        
+                plt.imshow(face_array)
+                st.pyplot(fig)
+
     elif choice == "Face Verification":
         st.subheader("Face Verification")
-        st.write("Please upload two image of the same person. The model will check if the two images contain the same face.")
-        st.write("Classifies whether the images match based on the probability score predicted by the model. If the difference is below the threshold (0.5 here) then the images are said to be identical.")
+        st.write(
+            "Please upload two image of the same person. The model will check if the two images contain the same face."
+        )
+        st.write(
+            "Classifies whether the images match based on the probability score predicted by the model. If the difference is below the threshold (0.5 here) then the images are said to be identical."
+        )
         column1, column2 = st.columns(2)
-    
+
         with column1:
-            image1 = st.file_uploader("Upload First Image", type=["jpg","png"], key="3")
-           
+            image1 = st.file_uploader(
+                "Upload First Image", type=["jpg", "png"], key="3"
+            )
+
         with column2:
-            image2 = st.file_uploader("Upload Second Image", type=["jpg","png"], key="4")
+            image2 = st.file_uploader(
+                "Upload Second Image", type=["jpg", "png"], key="4"
+            )
         # define filenames
-        if (image1 is not None) & (image2  is not None):
+        if (image1 is not None) & (image2 is not None):
             col1, col2 = st.columns(2)
-            image1 =  Image.open(image1)
-            image2 =  Image.open(image2)
+            image1 = Image.open(image1)
+            image2 = Image.open(image2)
             with col1:
                 st.image(image1)
             with col2:
                 st.image(image2)
 
-            filenames = [image1,image2]
-
+            filenames = [image1, image2]
             faces = [extract_face(f) for f in filenames]
-            # convert into an array of samples
-            samples = asarray(faces, "float32")
-            # prepare the face for the model, e.g. center pixels
-            samples = preprocess_input(samples, version=2)
-            # create a vggface model
-            model = VGGFace(model= "resnet50" , include_top=False, input_shape=(224, 224, 3),
-            pooling= "avg" )
-            # perform prediction
-            embeddings = model.predict(samples)
-            thresh = 0.5
-
-            score = cosine(embeddings[0], embeddings[1])
-            if score <= thresh:
-                st.success( " > Face is a match ( Score %.3f <= %.3f) " % (score, thresh))
+            print(faces)
+            faces_checked = [face for face in faces if face != []]
+            print(faces_checked)
+            if len(faces_checked) == 1:
+                st.error(
+                    "Could not find face in one of the images. Please Upload better quality images!"
+                )
+                st.stop()
             else:
-                st.error(" > Face is NOT a match ( Score %.3f > %.3f)" % (score, thresh))
+                # convert into an array of samples
+                samples = asarray(faces_checked, "float32")
+                # prepare the face for the model, e.g. center pixels
+                samples = preprocess_input(samples, version=2)
+                # create a vggface model
+                model = VGGFace(
+                    model="resnet50",
+                    include_top=False,
+                    input_shape=(224, 224, 3),
+                    pooling="avg",
+                )
+                # perform prediction
+                embeddings = model.predict(samples)
+                thresh = 0.5
+
+                score = cosine(embeddings[0], embeddings[1])
+                if score <= thresh:
+                    st.success(
+                        " > Face is a match ( Score %.3f <= %.3f) " % (score, thresh)
+                    )
+                else:
+                    st.error(
+                        " > Face is NOT a match ( Score %.3f > %.3f)" % (score, thresh)
+                    )
 
 
 def extract_face(file):
@@ -135,6 +172,8 @@ def extract_face(file):
     detector = MTCNN()
     # detect faces in the image
     results = detector.detect_faces(pixels)
+    if results == []:
+        return []
     # extract the bounding box from the first face
     x1, y1, width, height = results[0]["box"]
     x2, y2 = x1 + width, y1 + height
@@ -145,7 +184,6 @@ def extract_face(file):
     image = image.resize((224, 224))
     face_array = asarray(image)
     return face_array
-        
 
 
 if __name__ == "__main__":

requirements.txt

@@ -1,14 +1,11 @@
-Keras
-Keras-Applications
-Keras-Preprocessing
-git+https://github.com/ma7555/keras-vggface.git
-tensorflow-cpu
-mtcnn
+tensorflow
+streamlit
 matplotlib
-numpy
 oauthlib
-opencv-python-headless
 pandas
 Pillow
 scipy
-streamlit
+numpy==1.23.1
+protobuf==3.20.3
+mtcnn
+git+https://github.com/ma7555/keras-vggface.git