We aim to develop efficient and scalable algorithms to understand scenes in real-time from video stream(s).
In order to understand a scene, the relationship between different objects in the scene needs to be parsed.
Such relationships can be learned from existing videos using hierarchical graphical models (developed in earlier work funded by MOE and NRF), where more abstract concepts appear on a higher layer in the hierarchy. As concrete application, we will use these hierarchical models to detect complex events in time and space from single and multiple video streams.
More specifically, in this project we have the following aims:
• Development of smart algorithms for scene understanding
• Optimized and tested code and system, packaged into API. User document.
The research fellow will assist in conducting this research, writing journal publications, and preparing grant applications.
The candidate should have a PhD in Electrical Engineering, Industrial Engineering or Statistics
Required skills include:
• Signal processing
• Machine learning
• Data modelling
• Design of unmanned vehicles
Interested applicants may send his/her CV and supporting information to:
Asst Prof Justin Dauwels
School of Electrical & Electronic Engineering
Nanyang Technological University
50 Nanyang Avenue
E-mail Address for E-mailed Applications: JDAUWELS@ntu.edu.sg
Electronic submission of application is highly encouraged.
Only short-listed candidates will be notified for interview. Application closes when the positions are filled.