Update App (#1)

- Update app.py (09cbd669454d7d163b73102702af21efe0b6f411)

app.py

@@ -12,17 +12,26 @@ model_url = "https://huggingface.co/ElenaRyumina/face_emotion_recognition/resolv
 model_path = "FER_static_ResNet50_AffectNet.pth"
 
 response = requests.get(model_url, stream=True)
-with open(model_path, 'wb') as file:
+with open(model_path, "wb") as file:
     for chunk in response.iter_content(chunk_size=8192):
         file.write(chunk)
 
 pth_model = torch.jit.load(model_path)
 pth_model.eval()
 
-DICT_EMO = {0: 'Neutral', 1: 'Happiness', 2: 'Sadness', 3: 'Surprise', 4: 'Fear', 5: 'Disgust', 6: 'Anger'}
+DICT_EMO = {
+    0: "Neutral",
+    1: "Happiness",
+    2: "Sadness",
+    3: "Surprise",
+    4: "Fear",
+    5: "Disgust",
+    6: "Anger",
+}
 
 mp_face_mesh = mp.solutions.face_mesh
 
+
 def pth_processing(fp):
     class PreprocessInput(torch.nn.Module):
         def init(self):
@@ -37,24 +46,22 @@ def pth_processing(fp):
             return x
 
     def get_img_torch(img):
-        
-        ttransform = transforms.Compose([
-            transforms.PILToTensor(),
-            PreprocessInput()
-        ])
+        ttransform = transforms.Compose([transforms.PILToTensor(), PreprocessInput()])
         img = img.resize((224, 224), Image.Resampling.NEAREST)
         img = ttransform(img)
         img = torch.unsqueeze(img, 0)
         return img
+
     return get_img_torch(fp)
 
+
 def norm_coordinates(normalized_x, normalized_y, image_width, image_height):
-    
     x_px = min(math.floor(normalized_x * image_width), image_width - 1)
     y_px = min(math.floor(normalized_y * image_height), image_height - 1)
-    
+
     return x_px, y_px
 
+
 def get_box(fl, w, h):
     idx_to_coors = {}
     for idx, landmark in enumerate(fl.landmark):
@@ -63,44 +70,51 @@ def get_box(fl, w, h):
         if landmark_px:
             idx_to_coors[idx] = landmark_px
 
-    x_min = np.min(np.asarray(list(idx_to_coors.values()))[:,0])
-    y_min = np.min(np.asarray(list(idx_to_coors.values()))[:,1])
-    endX = np.max(np.asarray(list(idx_to_coors.values()))[:,0])
-    endY = np.max(np.asarray(list(idx_to_coors.values()))[:,1])
+    x_min = np.min(np.asarray(list(idx_to_coors.values()))[:, 0])
+    y_min = np.min(np.asarray(list(idx_to_coors.values()))[:, 1])
+    endX = np.max(np.asarray(list(idx_to_coors.values()))[:, 0])
+    endY = np.max(np.asarray(list(idx_to_coors.values()))[:, 1])
 
     (startX, startY) = (max(0, x_min), max(0, y_min))
     (endX, endY) = (min(w - 1, endX), min(h - 1, endY))
-    
+
     return startX, startY, endX, endY
 
-def predict(inp):
 
+def predict(inp):
     inp = np.array(inp)
     h, w = inp.shape[:2]
 
     with mp_face_mesh.FaceMesh(
-    max_num_faces=1,
-    refine_landmarks=False,
-    min_detection_confidence=0.5,
-    min_tracking_confidence=0.5) as face_mesh:
+        max_num_faces=1,
+        refine_landmarks=False,
+        min_detection_confidence=0.5,
+        min_tracking_confidence=0.5,
+    ) as face_mesh:
         results = face_mesh.process(inp)
         if results.multi_face_landmarks:
             for fl in results.multi_face_landmarks:
-                startX, startY, endX, endY  = get_box(fl, w, h)
-                cur_face = inp[startY:endY, startX: endX]
+                startX, startY, endX, endY = get_box(fl, w, h)
+                cur_face = inp[startY:endY, startX:endX]
                 cur_face_n = pth_processing(Image.fromarray(cur_face))
-                prediction = torch.nn.functional.softmax(pth_model(cur_face_n), dim=1).detach().numpy()[0]
+                prediction = (
+                    torch.nn.functional.softmax(pth_model(cur_face_n), dim=1)
+                    .detach()
+                    .numpy()[0]
+                )
                 confidences = {DICT_EMO[i]: float(prediction[i]) for i in range(7)}
-                
+
     return cur_face, confidences
 
+
 def clear():
     return (
         gr.Image(value=None, type="pil"),
-        gr.Image(value=None,scale=1, elem_classes="dl2"),
-        gr.Label(value=None,num_top_classes=3, scale=1, elem_classes="dl3")
+        gr.Image(value=None, scale=1, elem_classes="dl2"),
+        gr.Label(value=None, num_top_classes=3, scale=1, elem_classes="dl3"),
     )
 
+
 style = """
     div.dl1 div.upload-container {
         height: 350px;
@@ -154,26 +168,27 @@ with gr.Blocks(css=style) as demo:
                 submit = gr.Button(
                     value="Submit", interactive=True, scale=1, elem_classes="submit"
                 )
-                clear_btn = gr.Button(
-                    value="Clear", interactive=True, scale=1
-                )
+                clear_btn = gr.Button(value="Clear", interactive=True, scale=1)
         with gr.Column(scale=1, elem_classes="dl4"):
             output_image = gr.Image(scale=1, elem_classes="dl2")
             output_label = gr.Label(num_top_classes=3, scale=1, elem_classes="dl3")
     gr.Examples(
-        ["images/fig7.jpg", "images/fig1.jpg", "images/fig2.jpg","images/fig3.jpg",
-        "images/fig4.jpg", "images/fig5.jpg", "images/fig6.jpg"],
+        [
+            "images/fig7.jpg",
+            "images/fig1.jpg",
+            "images/fig2.jpg",
+            "images/fig3.jpg",
+            "images/fig4.jpg",
+            "images/fig5.jpg",
+            "images/fig6.jpg",
+        ],
         [input_image],
     )
-    
 
     submit.click(
         fn=predict,
         inputs=[input_image],
-        outputs=[
-            output_image,
-            output_label
-        ],
+        outputs=[output_image, output_label],
         queue=True,
     )
     clear_btn.click(
@@ -188,4 +203,4 @@ with gr.Blocks(css=style) as demo:
     )
 
 if __name__ == "__main__":
-    demo.queue(api_open=False).launch(share=False)
+    demo.queue(api_open=False).launch(share=False)