Update app.py

app.py

@@ -18,33 +18,34 @@ EMOTIONS = ['neutral', 'happiness', 'surprise', 'sadness', 'anger', 'disgust', '
 classifier = load_model(emotion_model_path)
 
 def predict_emotion(image):
+    # Convert the image to grayscale
+    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
 
-    faces = face_detection(image)
+    # Detect faces in the grayscale image
+    faces = face_detection.detectMultiScale(gray, scaleFactor=1.1,
+                                            minNeighbors=5, minSize=(30, 30),
+                                            flags=cv2.CASCADE_SCALE_IMAGE)
 
-    for face in faces:
-        x,y,w,h = face['box']
+    confidences = {}
+    for (x, y, w, h) in faces:
+        # Extract the region of interest (ROI)
+        roi_gray = gray[y:y+h, x:x+w]
+        roi_gray = cv2.resize(roi_gray, (48, 48), interpolation=cv2.INTER_AREA)
 
-        roi = image[y:y+h,x:x+w]
-        
-        # Converting the region of interest to grayscale, and resize
-        roi_gray = cv2.cvtColor(roi, cv2.COLOR_BGR2GRAY)
-        roi_gray = cv2.resize(roi_gray,(48,48),interpolation=cv2.INTER_AREA)
-
-        img = roi_gray.astype('float')/255.0
+        # Preprocess the ROI for prediction
+        img = roi_gray.astype('float') / 255.0
         img = img_to_array(img)
-        img = np.expand_dims(img,axis=0)
+        img = np.expand_dims(img, axis=0)
 
-        prediction = classifier.predict(img)[0]
-        #top_indices = np.argsort(prediction)[-2:]
-        #top_emotion = top_indices[1]
-        #second_emotion = top_indices[0]
-        #label = emotions[top_emotion]
-        confidences = {emotions[i]: float(prediction[i]) for i in range(len(emotions))}
+        # Make a prediction for the emotion
+        prediction = emotion_classifier.predict(img)[0]
+        confidences = {EMOTIONS[i]: float(prediction[i]) for i in range(len(EMOTIONS))}
 
     return confidences
 
 
 
+
 demo = gr.Interface(
     fn = predict_emotion,
     inputs = gr.Image(type="numpy"),