SkeletonVis: Interactive Visualization for Understanding Adversarial Attacks on Human Action Recognition Models

Haekyu Park, Zijie J. Wang, Nilaksh Das, Anindya S. Paul, Pruthvi Perumalla, Zhiyan Zhou, Duen Horng Chau

AAAI 2021 pp. 16094-16096

doi:10.1609/AAAI.V35I18.18022 /aaai/2021/park2021aaai-skeletonvis/

Abstract

Skeleton-based human action recognition technologies are increasingly used in video-based applications, such as home robotics, healthcare on the aging population, and surveillance. However, such models are vulnerable to adversarial attacks, raising serious concerns for their use in safety-critical applications. To develop an effective defense against attacks, it is essential to understand how such attacks mislead the pose detection models into making incorrect predictions. We present SkeletonVis, the first interactive system that visualizes how the attacks work on the models to enhance human understanding of attacks.

PDF AAAI Semantic Scholar

Cite

Text

Park et al. "SkeletonVis: Interactive Visualization for Understanding Adversarial Attacks on Human Action Recognition Models." AAAI Conference on Artificial Intelligence, 2021. doi:10.1609/AAAI.V35I18.18022

Markdown

[Park et al. "SkeletonVis: Interactive Visualization for Understanding Adversarial Attacks on Human Action Recognition Models." AAAI Conference on Artificial Intelligence, 2021.](https://mlanthology.org/aaai/2021/park2021aaai-skeletonvis/) doi:10.1609/AAAI.V35I18.18022

BibTeX

@inproceedings{park2021aaai-skeletonvis,
  title     = {{SkeletonVis: Interactive Visualization for Understanding Adversarial Attacks on Human Action Recognition Models}},
  author    = {Park, Haekyu and Wang, Zijie J. and Das, Nilaksh and Paul, Anindya S. and Perumalla, Pruthvi and Zhou, Zhiyan and Chau, Duen Horng},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2021},
  pages     = {16094-16096},
  doi       = {10.1609/AAAI.V35I18.18022},
  url       = {https://mlanthology.org/aaai/2021/park2021aaai-skeletonvis/}
}