Peng is currently an Asst. Prof. in ECE, AU. He was a Research Assoc. Prof. in USTC from 2021 to 2023. Before that, he was a PostDoc at CS, Univ. of Helsinki, where he got his PhD funded by EU Marie Curie ITN. His research interests focus on ensuring the trustworthiness of pervasive AI and XR systems, especially regarding efficiency, privacy, and fairness. He has publications on Ubicomp, Percom, ICPP, WWW, RecSys, ECCV, MM, VR, served as TPC for AAAI, IJCAI, MM, ISMAR, etc.
I focus on trustworthiness research in two directions, AI and XR, specifically regarding efficiency improvement, privacy protection, and fairness guaranteeing.
For TAI work, please refer to my recent publications as corresponding author:
Privacy-Preserving Orthogonal Aggregation for Guaranteeing Gender Fairness in Federated Recommendation, ACM WSDM 2025 (acceptance rate: 17.3%)
FedLoCA: Low-Rank Coordinated Adaptation with Knowledge Decoupling for Federated Recommendations, ACM RecSys 2024 (acceptance rate: 22%)
Heterogeneity-Aware Federated Deep Multi-View Clustering towards Diverse Feature Representations, ACM Multimedia 2024 (acceptance rate: 26%)
A + B: A General Generator-Reader Framework for Optimizing LLMs to Unleash Synergy Potential, ACL Findings 2024 (acceptance rate: 22.1%)
Check the full list on the topic.
For TXR work, please refer to my recent publications as corresponding author:
Unleashing GPT on the Metaverse: Savior or Destroyer? Proceedings of the IEEE (in press)
DreamScene: 3D Gaussian-based Text-to-3D Scene Generation via Formation Pattern Sampling, ECCV 2024
3D-GOI: 3D GAN Omni-Inversion for Multifaceted and Multi-object Editing, ECCV 2024
5G MEC Computation Handoff for Mobile Augmented Reality, IEEE MetaCom 2024
