First, clone the repository locally:
git clone https://github.com/wjf5203/VNext.git
cd VNext
Install dependencies and pycocotools for VIS:
pip install -r requirements.txt
pip install -e .
pip install shapely==1.7.1
pip install git+https://github.com/youtubevos/cocoapi.git#"egg=pycocotools&subdirectory=PythonAPI"
Compiling Deformable DETR CUDA operators:
cd projects/IDOL/idol/models/ops/
sh make.sh
Download and extract 2019 version of YoutubeVIS train and val images with annotations from CodeLab or YouTubeVIS, download OVIS and COCO 2017 datasets. Then, link datasets:
cd datasets/
ln -s /path_to_coco_dataset coco
ln -s /path_to_YTVIS19_dataset ytvis_2019
ln -s /path_to_ovis_dataset ovis
Extract YouTube-VIS 2019, OVIS, COCO 2017 datasets, we expect the directory structure to be the following:
VNext
├── datasets
│ ├──ytvis_2019
│ ├──ovis
│ ├──coco
...
ytvis_2019
├── train
├── val
├── annotations
│ ├── instances_train_sub.json
│ ├── instances_val_sub.json
...
ovis
├── train
├── valid
├── annotations_train.json
├── annotations_valid.jso
...
coco
├── train2017
├── val2017
├── annotations
│ ├── instances_train2017.json
│ ├── instances_val2017.json