Upload app.py

app.py

@@ -0,0 +1,54 @@
+from utils import create_model, get_optimal_font_scale
+from mtcnn import MTCNN
+import cv2
+import gradio as gr
+from PIL import Image
+import gradio
+
+model = create_model()
+checkpoint_path = "best_model_weights/checkpoint.ckpt"
+model.load_weights(checkpoint_path)
+
+detector = MTCNN()
+
+def face_mask_detection(image):
+    img = image.copy()
+    (h,w) = img.shape[:2]
+    objects = detector.detect_faces(img) # https://github.com/ipazc/mtcnn/blob/master/example.ipynb
+    
+    for obj in objects:
+        x, y, width, height = obj['box']
+        startX, startY, endX, endY = x, y, x+width, y+height
+
+        #ensure the bounding boxes fall within the dimensions of the frame
+        (startX,startY)=(max(0,startX),max(0,startY))
+        (endX,endY)=(min(w-1,endX), min(h-1,endY))
+
+
+        #extract the face ROI, convert it from BGR to RGB channel, resize it to 224,224 and preprocess it
+        face=img[startY:endY, startX:endX]
+        face=cv2.resize(face,(224,224))
+
+
+
+        probs = model.predict(face.reshape(1,224,224,3))[0][0] # example: result of model.predict: [[0.823] then < 0.5 -> mask, > 0.5 -> without mask
+        is_masked = probs < 0.5
+
+        #determine the class label and color we will use to draw the bounding box and text
+        label='Mask' if is_masked else 'No Mask'
+        color=(0,255,0) if label=='Mask' else (0,0,255)
+
+        #include the probability in the label
+        confidence = probs if probs > 0.5 else (1-probs)
+        label="{}: {:.2f}%".format(label,confidence*100)
+
+        #display the label and bounding boxes
+        fontScale = get_optimal_font_scale(label, endX-startX)
+        cv2.putText(img,label,(startX+5,startY+15),cv2.FONT_HERSHEY_SIMPLEX,fontScale,color,1)
+        cv2.rectangle(img,(startX,startY),(endX,endY),color,1)
+
+    return Image.fromarray(img)
+
+gr.Interface(fn=face_mask_detection,
+             inputs=gr.Image(shape=(224, 224)),
+             outputs=gr.outputs.Image(type = "pil", label = "Output Image")).launch()