Invited by: Prof. LIU Hao
AI Thrust, HKUST(GZ)
研讨会信息
🎤 Speaker: Prof. Flora Salim,
Professor,
Deputy Director of the UNSW AI Institute,
UNSW Sydney
📰 Title: Robust and Continual Learning for Streaming and Multimodal Data
⏰ Time: 3:00 PM,
📆 Date: Dec. 17, 2025 (Wednesday)
📍 Venue: E1-101, GZ Campus
研讨会概要
Abstract
Many enterprise and urban systems operate on continuous, heterogeneous data streams, from irregular sensor time series to large-scale videos. This talk presents recent advances toward building models that learn continually, adapt robustly, and generalize across such dynamic environments. I will discuss SeqLink and ODEStream, two frameworks for modeling partially observed and drifting time-series streams using Neural-ODE–based representations and buffer-free online adaptation. Extending to multimodal sensing, SensorLLM and ZARA explore how sensor data can be aligned with large language models for zero-shot recognition and interpretable reasoning. I will then cover several works on LLM post-tuning. One on reducing spurious correlations, and another on theory of mind. Finally, I will introduce ViLCo-Bench and Bisecle, which establish new benchmarks and methods for continual learning in video–language models through modality binding and task-aware separation. In summary, these works provide the building blocks towards adaptive, multimodal AI systems capable of operating reliably in dynamic real-world settings.
分享者简介
Prof. Flora Salim,
Flora Salim a full Professor at UNSW Sydney, where she also serves as a Deputy Director of the UNSW AI Institute. Her work focuses on multimodal machine learning and foundation models for time-series and spatio-temporal data, robust and trustworthy machine learning, and on AI and LLMs for smart and sustainable cities. She is a member of the Australian Academy of Sciences’ National Committee for Information and Computing Sciences and the Australian Research Council (ARC) College of Experts. She is a Vice Chair of the IEEE Task Force on AI for Time-Series and Spatio-Temporal Data. She serves in the editorial board of ACM TIST, ACM TSAS, PACM IMWUT, IEEE Pervasive Computing, Nature Scientific Data, and Machine Learning journal.
扫描加关注
获取更多AI学域消息
SCAN & FOLLOW US!

