Observation-Adaptive Spatiotemporal Positional Encoding: A State-Space Perspective
Tianyi Shi, Hanqun Cao, Yuhao Li, Pheng-Ann Heng, Weitao Du, Shengchao Liu.
Submitted to ICLR 2027.
Tianyi Shi 时天逸
MSc Student in Artificial Intelligence
The Chinese University of Hong Kong
About
I am a master's student in Artificial Intelligence at The Chinese University of Hong Kong (CUHK). I conduct research in the Wave Intelligence Lab under the supervision of Prof. Shengchao Liu. I received my bachelor's degree in Software Engineering from Tongji University in 2025.
My research focuses on multimodal learning and spatiotemporal perception. In September 2026, my work, "Observation-Adaptive Spatiotemporal Positional Encoding: A State-Space Perspective", received First Prize (250k) in AI for Research Algorithmic Problems at the Global Open-source AI Challenge (GOAI) 2026.
Research Interests
- Physical Intelligence
- Spatiotemporal Perception
- Multimodal Learning
- Representation Learning
Publications
MinkowskiPE: Minkowski Positional Encoding for Spatiotemporal Coupling
Yuhao Li, Louie Hong Yao, Tianyi Shi, Hanqun Cao, Hongxia Hao, Zhen Zhao, Shengchao Liu.
Submitted to ICLR 2027.
Learning to Optimize State Estimation in Multi-Agent Reinforcement Learning-Based Collaborative Detection
Tianlong Zhou, Tianyi Shi, Hongye Gao, Weixiong Rao.
IEEE Transactions on Mobile Computing (TMC), 2024.