Skip to content

kiyoon/video-adversarial-augmentation

Repository files navigation

Adversarial Augmentation Training Makes Action Recognition Models More Robust to Realistic Video Distribution Shifts

Kiyoon Kim, Shreyank N Gowda, Panagiotis Eustratiadis, Antreas Antoniou, Robert B Fisher
In ICPRAI 2024. arXiv springer

Adversarial Augmentation

Dataset downloads (matching classes only)

Running the code

pip install -e .
cp tools/run.env tools/.env
vi tools/.env  # change the settings here

# You can change the hyperparameters using environment variables like below
export ML_num_epochs=200
export ML_batch_size=16
export ML_lr_model=1e-4
# ... see more in BaseConfig in tools/run_aug.py

accelerate launch tools/run_aug.py

NOTE: you don't need to download the datasets as they will be downloaded automatically.

Citing the paper

If you find our work or code useful, please cite:

@InProceedings{kim2024videoadversarial,
author="Kim, Kiyoon
and Gowda, Shreyank N.
and Eustratiadis, Panagiotis
and Antoniou, Antreas
and Fisher, Robert B.",
editor="Wallraven, Christian
and Liu, Cheng-Lin
and Ross, Arun",
title="Adversarial Augmentation Training Makes Action Recognition Models More Robust to Realistic Video Distribution Shifts",
booktitle="Pattern Recognition and Artificial Intelligence",
year="2025",
publisher="Springer Nature Singapore",
address="Singapore",
pages="186--200",
isbn="978-981-97-8702-9"
}

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages