Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Label-Only Model Inversion Attacks via Boundary Repulsion

About

Recent studies show that the state-of-the-art deep neural networks are vulnerable to model inversion attacks, in which access to a model is abused to reconstruct private training data of any given target class. Existing attacks rely on having access to either the complete target model (whitebox) or the model's soft-labels (blackbox). However, no prior work has been done in the harder but more practical scenario, in which the attacker only has access to the model's predicted label, without a confidence measure. In this paper, we introduce an algorithm, Boundary-Repelling Model Inversion (BREP-MI), to invert private training data using only the target model's predicted labels. The key idea of our algorithm is to evaluate the model's predicted labels over a sphere and then estimate the direction to reach the target class's centroid. Using the example of face recognition, we show that the images reconstructed by BREP-MI successfully reproduce the semantics of the private training data for various datasets and target model architectures. We compare BREP-MI with the state-of-the-art whitebox and blackbox model inversion attacks and the results show that despite assuming less knowledge about the target model, BREP-MI outperforms the blackbox attack and achieves comparable results to the whitebox attack.

Mostafa Kahla, Si Chen, Hoang Anh Just, Ruoxi Jia• 2022

Related benchmarks

TaskDatasetResultRank
Model InversionCelebA (test)
Attack Accuracy75.67
36
Model Inversion AttackCelebA
Attack Acc73.93
16
Model Inversion AttackStandard MI attack datasets (CelebA, Pubfig83, FFHQ, Facescrub)
Queries (M)17.98
15
Model Inversion AttackPubfig83
Attack Accuracy72.8
8
Model Inversion AttackFacescrub
Attack Accuracy40.2
8
Model Inversion AttackPubfig83 (private identities)
Attack Accuracy0.66
4
Model Inversion AttackCelebA private identities
Attack Accuracy75.67
4
Model Inversion AttackFacescrub (private identities)
Attack Accuracy0.3568
4
Model Inversion AttackCelebA 128x128 resolution (test)
KNN Distance1.39e+3
4
Showing 9 of 9 rows

Other info

Follow for update