ML Systems Lab
Our lab studies the design and algorithms for robust and scalable ML systems, and the modelling and inference challenges arise when such systems interact with a diversity of entities in the physical world.
The work of our lab often falls into the following three themes. First, the development of new Machine Learning (ML) methods that enable systems to perceive and reason about different forms of information, such as images/videos, point clouds, trajectories and event streams. Second, the design of efficient algorithms that significantly reduce ML computation (from automated discovery/optimisation of ML architectures to model training and inference) with resource/hardware constraints in mind. Third, the applications of ML-powered analytics in managing and understanding data footprints generated by such systems across spatial and temporal domains. The overarching goal of our research is to build next-gen ML systems that are ubiquitous, evolving as they work alongside us to rely less on human assistance but more on computational intelligence, and finally become dependable and reliable.
We are based in the Department of Computer Science at the University of Warwick.
Current Members
Faculty
- Hongkai Wen
Research Staff
- Lichuan Xiang
- Konstantin Klemmer (EPSRC Mathematical Science Research Fellow)
Research Students
- Jiayang Gu
- Zhang Fang
- Shiwen Gan (visiting student from Nanjing U)
- Rosco Hunter
- Mo Wang (co-supervised with Quanying Liu @ SUSTech)
- Minghao Xu
- Hai Wang (visiting student from SEU, with Shuai Wang)
- Ajayi Olayinka (co-supervised with Tanaya Gupta @ Glasgow)
Interns
- Wenzhe Zhang (summer intern, now at Google)
- Tianyou Song (summer intern, now at Columbia U)
- Kun Li (summer intern, now at UIUC)
AutoCAML @ Samsung AI
We also work closely with the Automated (CAmbridge) Machine Learning (AutoCAML) group at Samsung AI Centre Cambridge (SAIC), which is part of the Embedded AI team at SAIC.
Joining
If you are interested in joining or collaborating with us please feel free to make contact. For perspective students you might also want to check the current openings.