Wednesday 13 February 2019 – Lunchtime seminar with Dr Kofi Appiah (Lecturer in Computer Science)
Targets of atypical size relative to their surrounding objects are usually missed by humans during visual search; studies show that this isn’t necessarily a deficiency but a useful strategy to discount distractors. These characteristics may explain why the human visual system works very well with processing of complex information in real time and can detect moving objects whilst in motion.
At present, the standard approach for detecting moving objects in a video scene can be achieved with either the unsupervised background subtraction approach with stationary cameras or the supervised state-of-the-art deep neural networks (DNN) with moving cameras. However, DNN guarantees the detection of only known objects with no sense of understanding the scene. The increasing popularity of using non-stationary cameras by several security agencies means that video data will have some elements of movement. Therefore, methods to reliably and robustly model video scenes with moving cameras are in great demand.
This talk will focus on the need to model objects in a scene with inspiration from physiological experiments.
Kofi Appiah received a Ph.D. in computer science in 2010, after working on various embedded computer vision projects as a research assistant at Aston University (Photonics Research Group), Durham University (Computer Vision Group) and University of Essex (Embedded and Intelligent Systems Group). He is currently a lecturer at Sheffield Hallam University.
Before joining Sheffield Hallam University in 2017, he worked at Nottingham Trent University as a lecturer from 2013. Kofi’s current research focuses on the interface between embedded computer vision, machine learning and neuroscience; mainly to understand and model the biological vision system on parallel architectures like Field Programmable Gate Arrays (FPGA).
He has published more than 50 peer-reviewed academic papers and serve as a reviewer for several core journals, such as Computer Vision and Image Understanding, IEEE Transactions on Neural Networks and Learning Systems, PLOS ONE, and more. In 2009, he won the IEEE Computational Intelligence Association Outstanding Student’s Award; in 2015, he secured a fully funded PhD studentship as a director of studies; and in 2017, he received the Sheffield Hallam University Higher Education Impact Research Fellowship to work with Sundance Ltd.
WEDNESDAY 13 FEBRUARY 2019
CANTOR 9138 , CITY CAMPUS, SHEFFIELD HALLAM UNIVERSITY
See here for details of other seminars in the series.
All SHU staff and students are welcome to attend the C3RI Lunchtime Research Seminars. If you are from outside of the University and would like to attend a seminar, please email the C3RI Administrator to arrange entry.