Summary of Adversary-guided Motion Retargeting For Skeleton Anonymization, by Thomas Carr et al.
Adversary-Guided Motion Retargeting for Skeleton Anonymization
by Thomas Carr, Depeng Xu, Aidong Lu
First submitted to arxiv on: 8 May 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A new anonymization technique is proposed in this paper for protecting user identity in virtual reality (VR) applications that utilize skeleton-based motion visualization. The method, called Privacy-centric Deep Motion Retargeting (PMR), uses adversarial learning to remove personally identifiable information (PII) embedded in the skeleton data. This is achieved by transferring the movement of the user onto a dummy skeleton, effectively linking any PII to the retargeted skeleton instead of the original user. The PMR model is designed to balance motion retargeting utility with privacy protection, achieving state-of-the-art performance on par with existing models while significantly reducing the effectiveness of privacy attacks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A new way to keep people’s identities private in virtual reality is being developed. In VR, people use special sensors to track their movements and create a digital skeleton. This skeleton can reveal who the person is, but a new technique called motion retargeting can change that. It moves the original movement onto a fake skeleton, making it harder for others to figure out who the person is. The goal is to keep the private information safe while still allowing people to have fun in VR. |
