#12 but only works when there's one face selected

Pytorch_Retinaface/pytorch_retinaface.py

@@ -106,7 +106,7 @@ class Pytorch_RetinaFace:
 
             # Crop the region and add padding to form a square
             cropped_imgs.append(image[crop_y1:crop_y2, crop_x1:crop_x2])
-            bbox_infos.append((original_crop_x1, original_crop_y1, original_crop_x2, original_crop_y2))
+            bbox_infos.append(((original_crop_x2-original_crop_x1, original_crop_y2-original_crop_y1),(original_crop_x1, original_crop_y1, original_crop_x2, original_crop_y2)))
         return cropped_imgs, bbox_infos
 
     def detect_faces(self, img):

__init__.py

@@ -132,13 +132,14 @@ class AutoCropFaces:
 
         # If we haven't selected anything, then return original images.
         if len(selected_faces) == 0: 
-            selected_crop_data = [(0, 0, img.shape[3], img.shape[2]) for img in original_images]
-            return (image, selected_crop_data)
+            # selected_crop_data = [(0, 0, img.shape[3], img.shape[2]) for img in original_images]
+            return (image, None)
 
         # If there is only one detected face in batch of images, just return that one.
         elif len(selected_faces) <= 1:
             out = selected_faces[0]
-            return (out, selected_crop_data)
+            crop_data = selected_crop_data[0] # to be compatible with WAS
+            return (out, crop_data)
 
         # Determine the index of the face with the maximum width
         max_width_index = max(range(len(selected_faces)), key=lambda i: selected_faces[i].shape[1])
@@ -165,6 +166,7 @@ class AutoCropFaces:
             else:
                 out = torch.cat((out, face_image), dim=0)
 
+        #TODO: WAS doesn't not support multiple faces, so this won't work with WAS.
         return (out, selected_crop_data)
 
 NODE_CLASS_MAPPINGS = {