Welcome to S3DI Lab!
The research lab for Secure, Scalable, and ReSponsible Distributed Intelligence (S3DI), led by Prof. Songze Li, works on a wide range of research topics to improve security, scalability, and trustworthiness of distributed computing frameworks, focusing on the applications of machine learning, artificial intelligence, and blockchain.
Our current research areas include:
1) Vulnerabilities and secure protocols for federated learning;
2) Security, privacy, and safety of large language and multi-modal models;
3) Secure multi-party computation;
4) Blockchain scalability and privacy.
We are always looking for strongly motivated PhD and Master students, Undergraduate students, Post-docs, Research assistants, and Visitors/interns to join our lab. Interested applicants please email your CV, transcript, and any related publications to songzeli [at] seu [dot] edu [dot] cn, or songzeli8824 [at] outlook [dot] com.
News
Oct 28, 2024 | Our paper “Generalized Lagrange Coded Computing: A Flexible Computation-Communication Tradeoff for Resilient, Secure, and Private Computation” is accepted to IEEE Transactions on Communications. |
---|---|
Oct 16, 2024 | Our paper “URVFL: Undetectable Data Reconstruction Attack on Vertical Federated Learning” is accepted to NDSS 2025. Congratulations to Duanyi! |
Sep 07, 2024 | Prof. Songze Li is serving Program Committee of NeurIPS 2024 International Workshop on Federated Foundation Models. |
Aug 05, 2024 | Our paper “Manifoldchain: Maximizing Blockchain Throughput via Bandwidth-Clustered Sharding” is accepted to NDSS 2025. Congratulations to Chunjiang! |
May 31, 2024 | Our paper “BackdoorIndicator: Leveraging OOD Data for Proactive Backdoor Detection in Federated Learning” is accepted to USENIX Security 2024. Congratulations to Yanbo! |
May 28, 2024 | Congratulations to Jin Liu on sucessfully defending his MPhil thesis! |
Mar 05, 2024 | Congratulations to Yanbo Dai on sucessfully defending his MPhil thesis! |
Jan 16, 2024 | Our paper “Constructing Adversarial Examples for Vertical Federated Learning: Optimal Client Corruption through Multi-Armed Bandit” is accepted to ICLR 2024. Congratulations to Duanyi! |