UoM Visual Intelligence Lab

Development of computer vision algorithms to detect objects, recognise scenes, and recreate 3D depth from images. The approach is inspired by human vision and involves modelling and learning the physical constraints of scenes to improve the accuracy, robustness, and interpretability of computer vision algorithms. Development of models of human visual attention to predict where people look in images, and investigate how people work with object detection algorithms in visual search tasks.

Facilities: A high-performance computing platform (Spartan), which offers a flexible infrastructure to support various computational needs, including a GPU partition with 1358 P100 Nvidia GPUs. Access to the Interaction Design Laboratory, a state-of-the-art usability lab, to run behavioural studies using psychophysics, eye tracking, and virtual reality. Dr Krista A. Ehinger is specifically affiliated with the Melbourne Centre for Data Science, an interdisciplinary centre focussed on big data analysis.

Contact: Dr Krista A. Ehinger, kris.ehinger@unimelb.edu.au, http://kehinger.com/