Modify flows and variable names

main.py

@@ -97,7 +97,7 @@ def filter_files(files: List) -> List:
     return file_paths
 
 
-def predict_bilateral(file):
+def predict_bilateral(files):
     """Predict Bilateral Mammography.
 
     Parameters
@@ -110,17 +110,23 @@ def predict_bilateral(file):
     List[List, Dict]
         List of objects that will be used to display the result
     """
+
+    filtered_files = filter_files(files)
+
     displays_imgs = []
+    images = []
 
-    image = np.array(Image.open(file.name))/257
-    image = np.reshape(image, (2, image.shape[0]//2, image.shape[1]))
+    for path in filtered_files:
+        image = np.array(Image.open(str(path)))
+        images.append(image)
 
-    im_h, im_w = image[0].shape[:2]
+    images = np.asarray(images).astype(np.float32)
+    im_h, im_w = images[0].shape[:2]
 
-    image_t = torch.from_numpy(image)
-    image_t = image_t.unsqueeze(0)  # Add batch dimension
+    images_t = torch.from_numpy(images)
+    images_t = images_t.unsqueeze(0)  # Add batch dimension
 
-    out, _, out_refiner = BILATERAL_MODEL(image_t)
+    out, _, out_refiner = BILATERAL_MODEL(images_t)
 
     out_refiner = utils.mean_activations(out_refiner).numpy()
 
@@ -135,10 +141,10 @@ def predict_bilateral(file):
         refined_view, (im_w, im_h), interpolation=cv2.INTER_LINEAR)
 
     image0_colored = cv2.normalize(
-        image[0], None, 0, 255, cv2.NORM_MINMAX, dtype=cv2.CV_8U)
+        images[0], None, 0, 255, cv2.NORM_MINMAX, dtype=cv2.CV_8U)
     image0_colored = cv2.cvtColor(image0_colored, cv2.COLOR_GRAY2RGB)
     image1_colored = cv2.normalize(
-        image[1], None, 0, 255, cv2.NORM_MINMAX, dtype=cv2.CV_8U)
+        images[1], None, 0, 255, cv2.NORM_MINMAX, dtype=cv2.CV_8U)
     image1_colored = cv2.cvtColor(image1_colored, cv2.COLOR_GRAY2RGB)
 
     heatmap0_overlay = cv2.addWeighted(