SoK: Data Reconstruction Attacks Against Machine Learning Models: Definition, Metrics, and Benchmark
This work addresses a foundational problem for researchers in machine learning security by providing standardized tools to advance the field, though it is incremental in building upon existing attack studies.
The paper tackles the lack of formal definitions and evaluation metrics for data reconstruction attacks in machine learning by proposing a unified taxonomy, quantitative metrics, and a benchmark framework, with empirical validation from a memorization perspective.
Data reconstruction attacks, which aim to recover the training dataset of a target model with limited access, have gained increasing attention in recent years. However, there is currently no consensus on a formal definition of data reconstruction attacks or appropriate evaluation metrics for measuring their quality. This lack of rigorous definitions and universal metrics has hindered further advancement in this field. In this paper, we address this issue in the vision domain by proposing a unified attack taxonomy and formal definitions of data reconstruction attacks. We first propose a set of quantitative evaluation metrics that consider important criteria such as quantifiability, consistency, precision, and diversity. Additionally, we leverage large language models (LLMs) as a substitute for human judgment, enabling visual evaluation with an emphasis on high-quality reconstructions. Using our proposed taxonomy and metrics, we present a unified framework for systematically evaluating the strengths and limitations of existing attacks and establishing a benchmark for future research. Empirical results, primarily from a memorization perspective, not only validate the effectiveness of our metrics but also offer valuable insights for designing new attacks.