Jingkang/PVSG

We carefully collect 400 videos, each featuring dynamic scenes and rich in logical reasoning content. On average, these videos are 76.5 seconds long (5 FPS). The collection comprises 289 videos from VidOR, 55 videos from EpicKitchen, and 56 videos from Ego4D.

Please git clone https://github.com/LilyDaytoy/OpenPVSG and organize your files according to the structure below.

You can put the HF dataset in data_zip directory.

├── assets
├── checkpoints
├── configs
├── data
├── data_zip
│   ├── Ego4D
│   │   ├── ego4d_masks.zip
│   │   └── ego4d_videos.zip
│   ├── EpicKitchen
│   │   ├── epic_kitchen_masks.zip
│   │   └── epic_kitchen_videos.zip
│   ├── VidOR
│   │   ├── vidor_masks.zip
│   │   └── vidor_videos.zip
│   └── pvsg.json
├── datasets
├── models
├── scripts
├── tools
├── utils
├── .gitignore
├── environment.yml
└── README.md

Please run unzip_and_extract.py to unzip the files and extract frames from the videos. If you use zip, make sure to use unzip -j xxx.zip to remove junk paths. You should have your data directory looks like this:

data
├── ego4d
│   ├── frames
│   ├── masks
│   └── videos
├── epic_kitchen
│   ├── frames
│   ├── masks
│   └── videos
├── vidor
│   ├── frames
│   ├── masks
│   └── videos
└── pvsg.json

We suggest our users to play with ./notebooks/Visualize_Dataset.ipynb to quickly get familiar with PVSG dataset.

Citation

For more information about the methods used in this dataset, please refer to the following paper:

Panoptic Video Scene Graph Generation, arXiv:2310.15166.