Junxiao Wang (王军晓)

alt text

Postdoctoral Fellow
Division of CEMSE
King Abdullah University of Science and Technology
Thuwal, Saudi Arabia
Work Email: email
Personal Email: email

Find me at  linkedin github google scholar

About Me

If you're interested in a research internship (remote), we welcome applicants of all levels (PhD/MSc/BSc). Don't hesitate to send me an email!

I currently hold the position of Postdoctoral Fellow at the KAUST Provable Responsible AI and Data Analytics (PRADA) Laboratory, working with Assistant Prof. Di Wang.

In 2021-2023, I was a Postdoctoral Fellow at the Pervasive Edge Intelligence Laboratory (PEIL) of The Hong Kong Polytechnic University, working with Prof. Song Guo. In 2018-2019, I was a Visiting Student at the Networks Research Group of Queen Mary University of London, working with Prof. Steve Uhlig. I completed my PhD at Dalian University of Technology in 2020, where my mentor was Prof. Keqiu Li. Prior to that, I earned my MEng and BE degrees in 2017, 2014.

Research Interests

I'm broadly interested in artificial intelligence and system with a special focus on distributed machine learning, AI security, privacy and explainability, and optimization of inference.

Recent Professional Activities

  • Reviewer, Asian Conference on Computer Vision (ACCV), 2024.01

  • Reviewer, European Conference on Computer Vision (ECCV), 2024.01

  • PC Member/Reviewer, International Conference on Artificial Intelligence and Statistics (AISTATS), 2023.09

  • Reviewer, International Conference on Computer Vision (ICCV), 2023.05

  • Postdoctoral Fellow, working with Assistant Prof. Di Wang, at Division of Computer, Electrical and Mathematical Sciences and Engineering (CEMSE), King Abdullah University of Science and Technology (KAUST), 2023.02

  • Invited Talk, Ritsumeikan University and CCF Dalian International Seminar, 2022.03

  • Postdoctoral Fellow, working with Prof. Song Guo, at Department of Computing (COMP), The Hong Kong Polytechnic University (PolyU), 2021.03-2023.02

  • Visiting Student, working with Prof. Steve Uhlig, at School of Electronic Engineering and Computer Science (EECS), Queen Mary University of London (QMUL), 2018.09-2019.09

News

  • Jan 2024   Our paper entitled "Faithful Vision-Language Interpretation via Concept Bottleneck Models" was accepted by ICLR2024.

  • Sep 2023   Our paper entitled "Towards Test-Time Refusals via Concept Negation" was accepted by NeurIPS2023.

  • Aug 2023   Our paper entitled "PromptFL: Let Federated Participants Cooperatively Learn Prompts Instead of Models — Federated Learning in Age of Foundation Model" was accepted by IEEE Transactions on Mobile Computing.

  • Jun 2023   Our paper entitled "On Knowledge Editing in Federated Learning: Perspectives, Challenges, and Future Directions" was available in arxiv.org.

  • May 2023   Our paper entitled "Investigating Trojan Attacks on Pre-trained Language Model-powered Database Middleware" was accepted by KDD2023.

  • Mar 2023   Our paper entitled "DualMix: Unleashing the Potential of Data Augmentation for Online Class-Incremental Learning" was available in arxiv.org.

  • Feb 2023   Our paper entitled "PMR: Prototypical Modal Rebalance for Multimodal Learning" was accepted by CVPR2023.

  • Jan 2023   Our paper entitled "pFedPrompt: Learning Personalized Prompt for Vision-Language Models in Federated Learning" was accepted by WWW2023.

  • Nov 2022   Our paper entitled "Demystify Self-Attention in Vision Transformers from a Semantic Perspective: Analysis and Application" was available in arxiv.org.

  • Nov 2022   Our paper entitled "PMR: Prototypical Modal Rebalance for Multimodal Learning" was available in arxiv.org.

  • Nov 2022   Our paper entitled "FedTune: A Deep Dive into Efficient Federated Fine-Tuning with Pre-trained Transformers" was available in arxiv.org.

  • Aug 2022   Our paper entitled "PromptFL: Let Federated Participants Cooperatively Learn Prompts Instead of Models — Federated Learning in Age of Foundation Model" was available in arxiv.org.

  • Aug 2022   Our paper entitled "Federated Unlearning: Guarantee the Right of Clients to Forget" was accepted by IEEE Network.

  • Apr 2022   Our paper entitled "Efficient Integrity Authentication Scheme for Large-scale RFID Systems" was accepted by IEEE Transactions on Mobile Computing.

  • Apr 2022   Our paper entitled "A Survey on Gradient Inversion: Attacks, Defenses and Future Directions" was accepted by IJCAI2022.

  • Feb 2022   Our paper entitled "Vertical Machine Unlearning: Selectively Removing Sensitive Information From Latent Feature Space" was available in arxiv.org.

  • Jan 2022   Our paper entitled "Federated Unlearning via Class-Discriminative Pruning" was accepted by WWW2022.

  • Dec 2021   Our paper entitled "Protect Privacy from Gradient Leakage Attack in Federated Learning" was accepted by INFOCOM2022.

  • Oct 2021   Our paper entitled "Federated Unlearning via Class-Discriminative Pruning" was available in arxiv.org.

  • Feb 2020   Our paper entitled "Click-UP: Toward the Software Upgrade of Click-Based Modular Network Function" was accepted by IEEE Systems Journal.

  • Jun 2018   Our paper entitled "CLICK-UP: Towards Software Upgrades of Click-driven Stateful Network Elements" was accepted by Demo track of SIGCOMM2018.