Date: Wednesday 8th January 2014
Place: Informatics Teaching Lab (ITL) — Top floor meeting room
Speaker: Xiatian(Eddy) Zhu
Title: Video Synopsis by Heterogeneous Multi-Source Correlation
Abstract: Generating coherent synopsis for surveillance video stream remains a formidable challenge due to the ambiguity and uncertainty inherent to visual observations. In contrast to existing video synopsis approaches that rely on visual cues alone, we propose a novel multi-source synopsis framework capable of correlating visual data and independent non-visual auxiliary information to better describe and summarise subtle physical events in complex scenes. Specifically, our unsupervised framework is capable of seamlessly uncovering latent correlations among heterogeneous types of data sources, despite the non-trivial heteroscedasticity and dimensionality discrepancy problems. Additionally, the proposed model is robust to partial or missing non-visual information. We demonstrate the effectiveness of our framework on two crowded public surveillance datasets.
Xiatian(Eddy) Zhu is currently a PhD student in our group and this presentation is about his recent publication in ICCV’13.