update

README.md

@@ -5,6 +5,29 @@ Pytorch implementation of our paper [Classification-Then-Grounding: Reformulatin
 We also won the 1st place of Video Relation Understanding (VRU) Grand Challenge in ACM Multimedia 2021, with a simplified version of our model.(The code for object tracklets generation is available at [here](https://github.com/Dawn-LX/VidVRD-tracklets))
 
 # Datasets
+Download the [ImageNet-VidVRD](https://xdshang.github.io/docs/imagenet-vidvrd.html) dataset and [VidOR](https://xdshang.github.io/docs/vidor.html) dataset, and put them in the following folder as
+
+```
+├── dataloaders
+│   ├── dataloader_vidvrd.py
+│   └── ...
+├── datasets
+│   ├── cache                       # cache file for our dataloaders
+│   ├── vidvrd-dataset
+│   │   ├── train
+│   │   ├── test
+│   │   └── videos
+│   ├── vidor-dataset
+│   │   ├── annotation
+│   │   └── videos
+│   └── GT_json_for_eval
+│       ├── VidORval_gts.json       # GT josn for evlauate generated by VidVRD-helper/prepare_gts_for_eval.py
+│       └── VidVRDtest_gts.json
+├── experiments   
+├── models
+├── ...
+
+```
 
 # Evaluation: