Evripidis Gkanias
I am a PhD student in Bio-robotics, supervised by Prof.
Barbara Webb at the Institute for Perception, Action and
Behaviour (IPAB), School of Informatics, the University of Edinburgh.
I am interested in neural representations and information theory. I currently try to understand and model
the underline mechanism of the mushroom bodies in the insect brain. My main scientific questions are:
how do the mushroom bodies embody a reinforcement learning mechanism; how do they create associations
among different modalities; and how is this used from other parts of the brain for memory acquisition
and retrieval.
My research interests include probabilistic machine learning (and reinforcement learning), neural
computation, bioinspired computer vision and multimodal integration.
Experience
Education
2018 - now ⚮ PhD in Bio-robotics, University of Edinburgh.
Subject: Memory acquisition and retrieval in the insect brain.
Supervisor: Prof. Barbara Webb
2015 - 2016 ⚮ MSc in Artificial Intelligence and Machine Learning, University of Edinburgh.
Thesis: Robocrab: data-driven adaptation of the evasion behaviour in fiddler crabs.
Supervisor: Prof. Barbara Webb
Grade: with Distinction.
2008 - 2013 ⚮ BSc in Computer Science and Information Systems, Aristotle University of Thessaloniki.
Thesis: Deep learning algorithms for multi-label data.
Supervisor: Dr. Grigorios Tsoumakas
Grade: First-class honours.
Professional
2017 - 2018 ⚮ Research Associate, Univeristy of Edinburgh.
I was responsible for investigating the information content of polarised light in relation to animal navigation - using machine learning and information theory - before
using the outcomes to develop a technical specification / design for manufacture of a novel robot sensor.
This task was part of the Invisible Cues project and it was a joined work of the University of Edinburgh and the
University of Sheffield
2016 - 2017 ⚮ Research Assistant, Univeristy of Edinburgh.
I focus on trying to imitate the learning mechanism of the larval Drosophila, which creates associations among odours and tastes. The goal is to create such a mechanism in neural level and put
it on a robot platform. The robot will try to find the gustatory source following the gradients of the associated odour.
This task was part of the Minimal project.
2014 - 2015 ⚮ Research Assistant, Centre for Research and Technology - Hellas (CERTH).
My main task was to implement a toolbox, using C# and the WPF subsystem, which could be used to analyse and compare human gestures, tracked using different capturing devices, i.e. Microsoft Kinect,
Vicon, WIMUs. I also implemented an extension of it, which was compatible with Unity3D.
This task was part of the ”RePlay” project.
Teaching
2019 - now ⚮ Teaching Supporter, Univeristy of Edinburgh.
I am tutoring, demonstrating and marking for a variety of courses including:
2013 - 2014 ⚮ Community Teaching Assistant, University of Illinois (coursera).
I was responsible for answering students' questions regarding the course material and assignments for the course
"Heterogeneous Parallel Programming" conducted by Prof. Wen-mei W. Hwu.
Publications
Journal articles
Schwarz, S., Clement, L., Gkanias, E., & Wystrach, A. (2020). How do backward-walking ants (Cataglyphis velox) cope with navigational uncertainty?. Animal Behaviour, 164, 133-142.
Link
Gkanias, E., Risse, B., Mangan, M., & Webb, B. (2019). From skylight input to behavioural output: a computational model of the insect polarised light compass. PLoS Comput Biol 15(7): e1007123.
Link
Conference proceedings
Stouraitis, T., Gkanias, E., Hemmi, J. M., & Webb, B. (2017, July). Predator Evasion by a Robocrab. In Conference on Biomimetic and Biohybrid Systems (pp. 428-439). Springer, Cham.
Link
Theses & Projects
Gkanias, E. (2016). Robocrab: data-driven adaptation of the evasion behaviour in fiddler crabs. Master Thesis, MSc in Artificial Intelligence and Machine Learning, University of Edinburgh.
pdf
Gkanias, E. (2013). Deep learning algorithms for multi-label data. Bachelor Thesis, BSc in Computer Science and Information Systems, Aristotle University of Thessaloniki.
pdf
Accomplishments
Funding & Awards
2019, June ⚮ The Percy Sladen Memorial Fund Grant.
Source: The Linnean Society.
2018, September ⚮ Robotics and Autonomous Systems Centre for Doctoral Training Studentship.
Source: Engineering and Physical Sciences Research Council (EPSRC).
2015, September ⚮ UK/EU Master's Scholarship.
Source: The University of Edinburgh.
Competitions
2016, November ⚮ Winning price for "Artificial Intelligence Driven Bots" challenge (Skype)
Source: Junction Hackathon.
Talks
Gkanias, E., Risse, B., Mangan, M., & Webb, B. (2019, November). From skylight input to behavioural output: a computational model of the insect polarised light compass. In International Navigation Conference, Edinburgh, UK.
slides
Stouraitis, T., Gkanias, E., Hemmi, J. M., & Webb, B. (2017, July). Predator Evasion by a Robocrab. In Conference on Biomimetic and Biohybrid Systems, Stanford University, CA.
slides
Posters
Gkanias, E., & Webb, B. (2019, October). Memory acquisition and retrieval in the insect brain. In Edinburgh Centre for Robotics fifth annual conference, Edinburgh, United Kingdom.
poster
Gkanias, E., Scaria, A., Vladis, N. A., Risse, B., Mangan, M., & Webb, B. (2019, August). Robustness of a model of the insects' celestial compass in realistic conditions. In International Conference on Invertebrate Vision, Bäckaskog Slott, Sweden.
poster
Gkanias, E., Lagogiannis, K., & Webb, B. (2018, October). Imitating the Drosophila Larval Learning Behaviour on a Robot. In Conference on Behavioral Neurogenetics of Drosophila Larva, the University of Edinburgh, United Kingdom.
poster
Pacella, D., Risse, B., Gkanias, E., Mangan, M., & Webb, B. (2018, July). Neural models of ant navigation in a realistic 3D world. In International Comference of Neuroethology, Brisbane, Australia.
poster
Work Address
Room 1.34Informatics Forum
10 Crichton Street
EH8 9AB
Edinburgh
United Kingdom
+44 (0) 131 650 6013