SAIL-Research

07/09/2026

SAIL members (KS Wong and Trần Dũng) are excited to attend ECCV 2026 to present our work.

H-SFP: Hierarchical Federated Learning with Decoupled Split-Model Prototyping
📅 Saturday, September 12, 2026
⏰ 3:30–5:30 PM
📍 Exhibition Hall, Poster #458

Please stop by and talk with them if you are interested in hierarchical federated learning, efficient and continual learning, model security, or any of our other research topics.

🫵 We also have several opportunities at VinUniversity for Vietnamese students and researchers seeking:
- Full-time research assistant positions
- Master’s and PhD opportunities
- Postdoctoral positions

🧭 Feel free to reach out to Prof. Kok-Seng before or during the conference (in person or via CVENT). He would be happy to discuss our research, potential collaborations, and opportunities to join SAIL (https://lnkd.in/gVZcc3ph).

22/08/2026

🚀 Postdoctoral Research Fellow Opening: Trustworthy AI (VinUni & UIUC) 🇻🇳🇺🇸

Are you a Vietnamese researcher with a completed PhD looking to make a global impact in Artificial Intelligence?

The Security and Artificial Intelligence Lab (https://www.sail-research.com/) is seeking a passionate Postdoctoral Researcher to join our team working on cutting-edge Trustworthy AI research!

Fellowship Details
📅 Duration: 2 Years total
🌍 Location Split: Year 1 at VinUniversity (Vietnam) & Year 2 at the University of Illinois Urbana-Champaign (UIUC) (USA)
💰 Compensation: Highly competitive remuneration package offered

What We Are Looking For
🎓 A Vietnamese national who has already obtained a PhD in Computer Science, AI, Machine Learning, Cybersecurity, or a related field.
📚 Strong publication record in top-tier AI/ML or security venues (e.g., NeurIPS, ICML, ICLR, CVPR, AAAI, IEEE S&P, ACM CCS).
🔬 Strong foundational background and research interests aligned with SAIL's focus on secure, robust, and trustworthy AI systems.

How to Apply: Interested candidates should send their Curriculum Vitae (CV), a brief statement of research interests, and representative publications directly to Kok-Seng Wong at [email protected].

08/08/2026

Excited to announce that our paper, “GARFIELD: Graph-Adaptive SSM for Explainable 3D Multi-Person WiFi Pose Estimation,” has been accepted to BMVC 2026! 🎉 The conference had an acceptance rate of 27.9%.

Camera-based human pose estimation can introduce significant privacy concerns, particularly in homes, workplaces, and healthcare settings. WiFi Channel State Information (CSI) provides a promising privacy-preserving alternative by sensing human movement through radio-frequency signals. However, existing approaches still face challenges in multi-person tracking, computational efficiency, and model interpretability.

In this work, we present GARFIELD, a State Space Model (SSM)-based framework designed to address these limitations. GARFIELD reduces the model size by 91.1%, requiring only 4.3M parameters compared with 48.2M in previous state-of-the-art approaches. We also introduce an explainable WiFi-CSI pose estimation approach that combines temporal and subcarrier saliency maps with Contrastive Query Attribution (CQA), providing greater insight into how the model makes its predictions.

31/05/2026

🎉 SAIL Research at CVPR 2026!

We are excited to announce that the Security and Artificial Intelligence Lab (SAIL) will present three papers at CVPR 2026 in Denver, Colorado, USA.

📍 Venue: Colorado Convention Center, Denver, Colorado, USA
🕒 Time zone: Denver time, MDT, UTC-6
📅 Date: Sunday, June 7, 2026

⭐ Main Technical Track

📌 Paper 1
HFedATM: Hierarchical Federated Domain Generalization via Optimal Transport and Regularized Mean Aggregation

👥 Authors: Thinh Nguyen, Trung Phan, Binh T. Nguyen, Khoa D Doan, Kok-Seng Wong

🕞 Session: Poster Session 6
⏰ Time: 3:30 P.M. - 5:30 P.M.
📍 Location: ExHall A

🔎 Findings Track

📌 Paper 2
Onboarding Without Forgetting: Hypernetwork Personalization with Data-Free Replay for Personalized Federated Learning

👥 Authors: Thinh Nguyen, Le Huy Khiem, Van-Tuan Tran, Khoa D Doan, Nitesh V Chawla, Kok-Seng Wong

🕢 Session: Findings Posters
⏰ Time: 7:30 A.M. - 9:00 A.M.
📍 Location: ExHall A

📌 Paper 3
Memory-efficient Continual Learning with Prototypical Exemplar Condensation

👥 Authors: Minh-Duong Nguyen, Thien-Thanh Dao, Le-Tuan Nguyen, Dung D. Le, Kok-Seng Wong

🕢 Session: Findings Posters
⏰ Time: 7:30 A.M. - 9:00 A.M.
📍 Location: ExHall A

👏 Congratulations to all authors and collaborators for these exciting contributions to federated learning, personalized federated learning, continual learning, and trustworthy AI.

Come meet us at CVPR 2026 and discuss our latest research! 🚀

23/10/2024

We are now hiring Research Assistants for an innovative project on Green Serverless Computing for AI at VinUniversity! 🌍💻
If you're passionate about AI, cloud computing, and sustainability, this is your chance to contribute to cutting-edge research while reducing the environmental impact of AI systems.

Position: Full-Time (Undergraduate RA positions also available)
Location: On-site at VinUniversity Campus
Duration: Initial contract of 6-12 months (extendable)
Project Focus: Energy-efficient AI training with serverless computing
Skills Needed: AI/ML basics, cloud platforms (AWS, Azure), Python programming, data analysis

[More on Job Descriptions]: https://drive.google.com/file/d/1zTdqJJkBVw_z4UZ11V2T3IKkfabvezVZ/view?usp=drive_link

📄 How to Apply: Send your CV and cover letter to [email protected] and [email protected].

06/06/2024

As the time is closing in, we're very nervous but also feeling very proud of the team and the collaborators of our Farm2Vet solution to reduce Antimicrobial Resistance (AMR) with Generative AI (read about the solution here https://solve.mit.edu/.../trinity.../solutions/81996 and the official announcement here https://www.linkedin.com/.../the-trinity-challenge... ).
A pat on the shoulder to MAIL-Research and SAIL-Research (and of course everyone else on the solution). Very luck that we have a chance to collaborate with everyone!!!

From our Founder and Chair, Dame Sally Davies, our Director, Prof Marc Mendelson and all the team - we want to wish all our finalists good luck for this afternoon!
Your innovative thinking and commitment is key in mitigating the .

06/06/2024

From our Founder and Chair, Dame Sally Davies, our Director, Prof Marc Mendelson and all the team - we want to wish all our finalists good luck for this afternoon!
Your innovative thinking and commitment is key in mitigating the .

06/10/2023

Contact us in case you cannot find your "front door".....😁

Federated learning (FL) is an approach within the realm of machine learning (ML) that allows the use of distributed data without compromising personal…