yuxindu
/

segvol

Image Segmentation

feature-extraction

Model card Files Files and versions Community

yuxin commited on Apr 15

Commit

ddf6466

•

1 Parent(s): d0601bd

add config

Files changed (1) hide show

model_segvol_single.py +1 -1

model_segvol_single.py CHANGED Viewed

@@ -520,7 +520,7 @@ class SegVol(nn.Module):
             # organ prompt
             iter_organs.append(training_organs)
             # box prompt
-            box = generate_box(train_labels[sample_idx])
             iter_bboxes.append(box)
             # point prompt
             num_positive_extra_max, num_negative_extra_max = 10, 10

             # organ prompt
             iter_organs.append(training_organs)
             # box prompt
+            box = generate_box(train_labels[sample_idx], bbox_shift=10)
             iter_bboxes.append(box)
             # point prompt
             num_positive_extra_max, num_negative_extra_max = 10, 10