diff --git a/README.md b/README.md
index 4729d8b..bd71ddd 100644
--- a/README.md
+++ b/README.md
@@ -33,7 +33,7 @@ Then update the train/val/test list paths in the config files.
 + The train/val lists for PASCAL5i contain 5953 and 1449 images respectively. The train list should be **voc_sbd_merge_noduplicate.txt** and the val list is the original val list of pascal voc (**val_fixed.txt**).
 
 ##### To get voc_sbd_merge_noduplicate.txt:
-+ we first merge the original VOC (voc_original_train.txt) and SBD ([**sbd_data.txt**](http://home.bharathh.info/pubs/codes/SBD/train_noval.txt)) training data. 
++ We first merge the original VOC (voc_original_train.txt) and SBD ([**sbd_data.txt**](http://home.bharathh.info/pubs/codes/SBD/train_noval.txt)) training data. 
 + The merged list (voc_sbd_merge.txt) is then processed by the script (duplicate_removal.py) to remove the duplicate images and labels.
 
 ### Run Demo / Test with Pretrained Models