Liangli Zhen

Senior Scientist at A*STAR

zhenll.png

Welcome to my homepage! I am a group leader and senior scientist at the Institute of High Performance Computing, Agency for Science, Technology and Research (A*STAR), Singapore, and a Senior Member of the IEEE. I lead the AI Safety Research Group focused on enhancing AI safety and robustness for real-world deployment and application. Prior to joining A*STAR, I received my PhD in Computer Science at Sichuan University in 2018, under the supervision of Professor Dezhong Peng. I was a joint PhD student at The University of Birmingham, where I was supervised by Professor Xin Yao.

My research is mainly on machine learning and optimisation. Related research topics include adversarial attacks, domain generalisation, jailbreak attacks, hallucination mitigation, multimodal learning, multi-objective optimisation, and more.

News

  • [Jan 2026] We currently have openings for AI Safety research scientists.
  • [Jan 2026] Our paper on Safeguarding Federated Learning from Data Reconstruction Attacks via Gradient Dropout has been accepted for publication by IEEE Transactions on Information Forensics and Security.
  • [Jan 2026] Our paper on Noise-Adaptive Diffusion Sampling for Inverse Problems Without Task-Specific Tuning has been accepted for publication by ICLR 2026.
  • [Jan 2026] I will be serving as an Associate Editor for IEEE Transactions on Neural Networks and Learning Systems.
  • [Dec 2025] We have openings for visiting PhD students (including CSC Scholarship) and ARAP scholarship applicants.
  • [Sept 2025] Our paper on Single-Domain Generalisation via Path Flatness-Aware Optimisation of Loss Landscapes has been accepted for publication by IEEE Transactions on Neural Networks and Learning Systems.
  • [Aug 2025] Our paper on GAPNet: A Lightweight Framework for Image and Video Salient Object Detection via Granularity-Aware Paradigm has been accepted for publication by Machine Intelligence Research.
  • [Jun 2025] Our paper on Low-Resolution Self-Attention for Semantic Segmentation has been accepted for publication by IEEE Transactions on Pattern Analysis and Machine Intelligence.
  • [May 2025] 🏆 Our team ranked the 3rd Place in the Global Challenge for Safe and Secure LLMs (Track 2: Defence).
  • [May 2025] Our project on Development of Stable, Robust and Secure Intelligent Systems for Autonomous Vehicles (Stage 1b) has been awarded the AI Singapore Robust AI Grand Challenge grant funding.
  • [Apr 2025] Our project on Advanced Optical Metrology for Asian Skin Health and Diseases (OPTIMAS) has been awarded the RIE2025 Industry Alignment Fund Pre-positioning (IAF-PP) grant funding.
  • [Apr 2025] Our paper on A Survey and Evaluation of Adversarial Attacks in Object Detection has been accepted for publication by IEEE Transactions on Neural Networks and Learning Systems.
  • [Dec 2024] Our paper on Deep Evidential Hashing for Trustworthy Cross-Modal Retrieval has been accepted for publication by the AAAI Conference on Artificial Intelligence 2025.
  • [Aug 2024] Our project on Cyber Sentinel: A Unified Framework for Safeguarding Foundation Models has been awarded the CyberSG Research and Development Programme Office (CRPO) Grand Challenge grant funding.
  • [Aug 2024] Our paper on Continuous Disentangled Joint Space Learning for Domain Generalisation has been accepted for publication by IEEE Transactions on Neural Networks and Learning Systems.
  • [May 2024] 🏆 Our team won the 1st Place in the IJCAI Vision-based Remote Physiological Signal Sensing Challenge.
  • [Mar 2024] Our paper on Generative Image Reconstruction from Gradients has been accepted for publication by IEEE Transactions on Neural Networks and Learning Systems.
  • [Jan 2024] Our paper on Geometric Correspondence-Based Multimodal Learning for Ophthalmic Image Analysis has been accepted for publication by IEEE Transactions on Medical Imaging.
  • [Jan 2024] Our paper on MedNAS: Multiscale Training-Free Neural Architecture Search for Medical Image Analysis has been accepted for publication by IEEE Transactions on Evolutionary Computation.
  • [Nov 2023] Our paper on Deep Supervised Multi-View Learning with Graph Priors has been accepted for publication by IEEE Transactions on Image Processing.
  • [Jul 2023] Our project on Development of Stable, Robust and Secure Intelligent Systems for Autonomous Vehicles (Stage 1a) has been awarded the AI Singapore Robust AI Grand Challenge grant funding.