Update data_format and label_list spec

README.md

@@ -128,7 +128,7 @@ config/datasets_mixed_odvg.json      # support mixed dataset for both OD and VG
 # Training
 
 - Before starting the training, you need to modify the  ``config/datasets_vg_example.json`` according to ``data_format.md``.
-- The evaluation code defaults to using coco_val2017 for evaluation. If you are evaluating with your own test set, you need to convert the test data to coco format (not the ovdg format in data_format.md), and modify the config to set **use_coco_eval = False** (The COCO dataset has 80 classes used for training but 90 categories in total, so there is a built-in mapping in the code).
+- The evaluation code defaults to using coco_val2017 for evaluation. If you are evaluating with your own test set, you need to convert the test data to coco format (not the ovdg format in data_format.md), and modify the config to set **use_coco_eval = False** (The COCO dataset has 80 classes used for training but 90 categories in total, so there is a built-in mapping in the code). Also, update the **label_list** in the config with your own class names like **label_list=['dog', 'cat', 'person']**.
 
 
 ```  bash

data_format.md

@@ -51,7 +51,7 @@ Visual Grounding datasets employ the 'grounding' field.
   "detection": {
       "instances": [     
         {
-          "bbox": [262,210,323,338],
+          "bbox": [262,210,323,338],   # [x1,y1,x2,y2]
           "label": 0,
           "category": "dog"
         },
@@ -71,7 +71,7 @@ Visual Grounding datasets employ the 'grounding' field.
       "caption": "a wire hanger with a paper cover that reads we heart our customers", 
       "regions": [
         {
-          "bbox": [20,215,985,665],
+          "bbox": [20,215,985,665],   # [x1,y1,x2,y2]
           "phrase": "a paper cover that reads we heart our customers"
         },
         {