Bilkent University
Department of Computer Engineering

---

CS 590/690 SEMINAR

 

CURE: Privacy-Preserving Split Learning Done Right

 

Aqsa Shabbir
Master Student
(Supervisor: Asst.Prof.Sinem Sav)
Computer Engineering Department
Bilkent University

Abstract: Training deep neural networks often requires large-scale datasets, necessitating storage and processing on cloud servers due to computational constraints. The procedures must follow strict privacy regulations in domains like finance and healthcare. Split Learning (SL), a framework that divides model layers between client(s) and server(s), is widely adopted for distributed model training. While Split Learning reduces privacy risks by limiting server access to the full parameter set, previous research has identified that intermediate outputs exchanged between server and client can compromise the client's data privacy. Homomorphic encryption (HE)-based solutions exist for this scenario but often impose prohibitive computational burdens. To address these challenges, we propose a novel system, based on HE, that encrypts the server side of the model parameters. Through optimized task allocation and computation strategies, this integration not only ensures end-to-end security but also minimizes computational overhead.

 

DATE: April 14, Thursday @ 14:40 Place: EA 502