upd readme

README.md

 ## GDRNPP for BOP2022
 
-This repo provides code and models for GDRNPP.
+This repo provides code and models for GDRNPP_BOP2022.
 
-## Path setting
+TODO: add authors
 
+## Path Setting
+
+### Dataset Preparation
+Download the 6D pose datasets from the
+[BOP website](https://bop.felk.cvut.cz/datasets/) and
+[VOC 2012](https://pjreddie.com/projects/pascal-voc-dataset-mirror/)
+for background images.
+Please also download the `image_sets` and `test_bboxes` from
+here [OneDrive](#TODO).
+
+The structure of `datasets` folder should look like below:
 ```
-# recommend using soft links (ln -sf)
 datasets/
 ├── BOP_DATASETS   # https://bop.felk.cvut.cz/datasets/
     ├──lm
@@ -15,12 +25,23 @@ datasets/
     ├──hb
     ├──itodd
     ├──tless
-├── VOCdevkit
-└── coco
+└──VOCdevkit
 ```
 
-## Dependencies
-See [INSTALL.md](./docs/INSTALL.md)
+
+### MODELS
+
+Download the trained models at [Onedrive](https://mailstsinghuaeducn-my.sharepoint.com/:f:/g/personal/liuxy21_mails_tsinghua_edu_cn/EgOQzGZn9A5DlaQhgpTtHBwBGWEB57mpYy4SbmpZJMmMyQ?e=0z9Xd2) and put them in the folder ./output.
+
+
+## Requirements
+* Ubuntu 16.04/18.04, CUDA 10.1/10.2, python >= 3.6, PyTorch >= 1.6, torchvision
+* Install `detectron2` from [source](https://github.com/facebookresearch/detectron2)
+* `sh scripts/install_deps.sh`
+* Compile the cpp extension for `farthest points sampling (fps)`:
+    ```
+    sh core/csrc/compile.sh
+    ```
 
 ## Detection
 TODO: tjw
@@ -41,8 +62,5 @@ The difference between this repo and gdrn conference version mainly including:
 
 `./core/gdrn_modeling/test_gdrn.sh <config_path> <gpu_ids> <ckpt_path> (other args)`
 
-The trained models can be found at [Onedrive](https://mailstsinghuaeducn-my.sharepoint.com/:f:/g/personal/liuxy21_mails_tsinghua_edu_cn/EgOQzGZn9A5DlaQhgpTtHBwBGWEB57mpYy4SbmpZJMmMyQ?e=0z9Xd2)
-
-
 ## Pose Refinement
 TODO: rudy
\ No newline at end of file