We are delighted to announce that Prof. YANG Qiang, our Affiliated Chair Professor of Artificial Intelligence, and Prof. LIU Yang Veronica, our Associate Professor, together with their co-authors from various institutions, have received the Runner Up for the Best Paper in the Dataset and Benchmark Track at the prestigious ACM SIGKDD 2025 Conference on Knowledge Discovery and Data Mining, held in Toronto, Canada, from 3-7 August, 2025.
About the Winning Paper:
Title: HtFLlib: A Comprehensive Heterogeneous Federated Learning Library and Benchmark
Authors: Jianqing Zhang (Shanghai Jiao Tong University), Xinghao Wu (Peking University), Yanbing Zhou (Chongqing University), Xiaoting Sun (Tongji University), Qiqi Cai (Shanghai Jiao Tong University), Yang Liu (The Hong Kong Polytechnic University), Yang Hua (Queen’s University Belfast), Zhenzhe Zheng (Shanghai Jiao Tong University), Jian Cao (Shanghai Jiao Tong University), Qiang Yang (The Hong Kong Polytechnic University)
As AI developing for many years, institutions already have built their specific models with heterogeneous model architectures for specific local tasks. Collaboration among heterogeneous models helps overcome data scarcity by enabling knowledge transfer across institutions and devices. Heterogeneous Federated Learning (HtFL) has emerged as a rapidly growing research area that allows participants to collaborate using their heterogeneous models, broadening the scope of traditional FL and fostering wider participation. This paper introduces the first Heterogeneous Federated Learning Library (HtFLlib), an easy-to-use and extensible framework that integrates multiple datasets and model heterogeneity scenarios, featuring 40 heterogeneous model architectures and 12 datasets and offering a robust benchmark for research and practical applications.
Click here to read the paper.
Click here to know more about the project.
About KDD
The ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) is the premier international forum for data science, machine learning, and AI researchers and practitioners from academia, industry, and government. It provides a platform to share cutting-edge research, innovative applications, and real-world experiences.
