About Me
I am a Senior Research Scientist at Institute for Infocomm Research, A*STAR. Previously, I was a postdoctoral researcher at UCL, supervised by Prof. Massimiliano Pontil and an applied scientist intern at AWS Berlin, mentored by Dr. Cedric Archambeau.
I earned my Ph.D. in Machine Learning from Imperial College London, with supervision from Prof. Yiannis Demiris and Prof. Carlo Ciliberto, and was supported by the Singapore National Science Scholarship.
Research Interests
My research spans a range of machine learning topics, including representation learning, meta-learning, continual learning, and imitation learning. My primary aim is to develop robust ML systems that efficiently leverage existing model knowledge to address novel tasks and are capable of continuous improvements.
Publications
- Schedule-Robust Continual Learning, R Wang, M Ciccone, M Pontil, C Ciliberto, under review for PAMI
- Robust Meta-Representation Learning via Global Label Inference and Classification, R Wang, I Falk, M Pontil, C Ciliberto, PAMI 2023
- Investigating Vision Foundational Models for Tactile Representation Learning, B Zandonati, R Wang, R Gao, Y Wu, arXiv 2022
- The Role of Global Labels in Few-Shot Classification and How to Infer Them, R Wang, M Pontil, C Ciliberto, NeurIPS 2021
- Structured prediction for conditional meta-learning, R Wang, Y Demiris, C Ciliberto, NeurIPS 2020
- Random expert distillation: Imitation learning via expert policy support estimation, R Wang, C Ciliberto, P Amadori, Y Demiris, ICML 2019
- Support-weighted adversarial imitation learning, R Wang, C Ciliberto, P Amadori, Y Demiris, Neurips LIRE Workshop 2019
- Real-time workload classification during driving using hypernetworks, R Wang, PV Amadori, Y Demiris, IROS 2018
- Multi-modal robot apprenticeship: imitation learning using linearly decayed dmp+ in a human-robot dialogue system, Y Wu, R Wang, LF D’Haro, RE Banchs, KP Tee, IROS 2018
- Magan: Margin adaptation for generative adversarial networks, R Wang, A Cully, HJ Chang, Y Demiris, arXiv 2017
- Dynamic movement primitives plus: For enhanced reproduction quality and efficient trajectory modification using truncated kernels and local biases, R Wang, Y Wu, WL Chan, KP Tee, IROS 2016