LetheSec/PLG-MI-Attack
[AAAI 2023] Pseudo Label-Guided Model Inversion Attack via Conditional Generative Adversarial Network
Python
Issues
- 0
Isn't training the cGAN with a private dataset already considered a privacy leak?
#2 opened by lym-123-123 - 0
Isn't training the cGAN with a private dataset already considered a privacy leak?
#3 opened by lym-123-123 - 1
cause the generator of KEDMI and GMI are not cGAN，how can I konw the recovery result of specific class?Or I should use the target model to select by myself?
#1 opened by nevereverforget1