Ho Kei Cheng, Yu-Wing Tai, Chi-Keung Tang

CVPR 2021

1. python download_model.py to get all the required models.
2. python interactive_gui.py --video <path to video> or python interactive_gui.py --images <path to a folder of images>. A video has been prepared for you at examples/example.mp4.
3. If you need to label more than one object, additionally specify --num_objects <number_of_objects>. See all the argument options with python interactive_gui.py --help.
4. There are instructions in the GUI. You can also watch the demo videos for some ideas.

See eval_interactive_davis.py. If you have downloaded the datasets and pretrained models using our script, you only need to specify the output path, i.e., python eval_interactive_davis.py --output [somewhere].

Go to this repo: STCN.

(Performance mostly bottlenecked by interaction. Speed-wise it improves quite a bit -- try the GUI.)

python download_model.py should get you all the models that you need. (pip install gdown required.)

Google Drive:

[stcn.pth] [fusion_stcn.pth]

OneDrive:

[stcn.pth] [fusion_stcn.pth]

Use the same commands as in master.

Please cite our papers if you find this repo useful!

@inproceedings{cheng2021stcn,
  title={Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation},
  author={Cheng, Ho Kei and Tai, Yu-Wing and Tang, Chi-Keung},
  booktitle={NeurIPS},
  year={2021}
}
@inproceedings{cheng2021mivos,
  title={Modular Interactive Video Object Segmentation: Interaction-to-Mask, Propagation and Difference-Aware Fusion},
  author={Cheng, Ho Kei and Tai, Yu-Wing and Tang, Chi-Keung},
  booktitle={CVPR},
  year={2021}
}

Contact: hkchengrex@gmail.com