Skip to main navigation Skip to search Skip to main content

Adversarial VAE Sanitization for Privacy-Aware Data Sharing in Semiconductor Manufacturing

  • Luis F. Garcia
  • , Stefano Rini*
  • , Shun C. Chang
  • , Ying C. Hsu
  • *Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

1 Scopus citations

Abstract

Semiconductor manufacturing data sharing is constrained by intellectual property (IP) concerns, which makes it difficult to apply collaborative learning even when multiple installations of the same equipment exist. This paper studies the applicability of adversarial variational autoencoders (VAE) as a practical sanitization mechanism for this setting. Sensitive process variables are specified by a domain expert, and the model is trained to preserve utility while making those variables harder to infer from the released representation. We evaluate the approach on native semiconductor manufacturing time-series runs, first in the original single-party setting and then in a small multi-peer extension with local and joint adversaries. The results suggest a privacy-utility trade-off: sanitized representations retain the main public operating regimes while degrading reconstruction of the protected variables. The distributed study further shows that leakage must be evaluated after aggregation, not only per peer. These results position adversarial VAE sanitization as a useful applied tool for semiconductor data sharing, while also highlighting the limitations of the current centralized simulation and small number of available runs.

Original languageEnglish
Pages (from-to)381-387
Number of pages7
JournalIEEE Transactions on Semiconductor Manufacturing
Volume39
Issue number3
DOIs
StatePublished - 1 Aug 2026

Keywords

  • Data obfuscation
  • IP-protection
  • privacy
  • semiconductor manufacturing
  • variational autoencoders

Fingerprint

Dive into the research topics of 'Adversarial VAE Sanitization for Privacy-Aware Data Sharing in Semiconductor Manufacturing'. Together they form a unique fingerprint.

Cite this