Unmasking Puppeteers: Leveraging Biometric Leakage to Expose Impersonation in AI-Based Videoconferencing
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909038147534848 |
|---|---|
| author | Vahdati, Danial Samadi Nguyen, Tai Duc Prashnani, Ekta Nagano, Koki Luebke, David Gallo, Orazio Stamm, Matthew |
| author_facet | Vahdati, Danial Samadi Nguyen, Tai Duc Prashnani, Ekta Nagano, Koki Luebke, David Gallo, Orazio Stamm, Matthew |
| contents | AI-based talking-head videoconferencing systems reduce bandwidth by sending a compact pose-expression latent and re-synthesizing RGB at the receiver, but this latent can be puppeteered, letting an attacker hijack a victim's likeness in real time. Because every frame is synthetic, deepfake and synthetic video detectors fail outright. To address this security problem, we exploit a key observation: the pose-expression latent inherently contains biometric information of the driving identity. Therefore, we introduce the first biometric leakage defense without ever looking at the reconstructed RGB video: a pose-conditioned, large-margin contrastive encoder that isolates persistent identity cues inside the transmitted latent while cancelling transient pose and expression. A simple cosine test on this disentangled embedding flags illicit identity swaps as the video is rendered. Our experiments on multiple talking-head generation models show that our method consistently outperforms existing puppeteering defenses, operates in real-time, and shows strong generalization to out-of-distribution scenarios. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_03548 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Unmasking Puppeteers: Leveraging Biometric Leakage to Expose Impersonation in AI-Based Videoconferencing Vahdati, Danial Samadi Nguyen, Tai Duc Prashnani, Ekta Nagano, Koki Luebke, David Gallo, Orazio Stamm, Matthew Computer Vision and Pattern Recognition Artificial Intelligence AI-based talking-head videoconferencing systems reduce bandwidth by sending a compact pose-expression latent and re-synthesizing RGB at the receiver, but this latent can be puppeteered, letting an attacker hijack a victim's likeness in real time. Because every frame is synthetic, deepfake and synthetic video detectors fail outright. To address this security problem, we exploit a key observation: the pose-expression latent inherently contains biometric information of the driving identity. Therefore, we introduce the first biometric leakage defense without ever looking at the reconstructed RGB video: a pose-conditioned, large-margin contrastive encoder that isolates persistent identity cues inside the transmitted latent while cancelling transient pose and expression. A simple cosine test on this disentangled embedding flags illicit identity swaps as the video is rendered. Our experiments on multiple talking-head generation models show that our method consistently outperforms existing puppeteering defenses, operates in real-time, and shows strong generalization to out-of-distribution scenarios. |
| title | Unmasking Puppeteers: Leveraging Biometric Leakage to Expose Impersonation in AI-Based Videoconferencing |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2510.03548 |