
Kosmas Deligkaris, PhD
Deep Learning and Machine Vision Engineer
My current work focuses on developing methods and tools for applied environmental AI, mostly related to deep-learning. You can find me at the Okinawa Institute of Science and Technology, a magnificent graduate university located in the beautiful island of Okinawa.
The methods and tools I envision have a wide-range of applications in machine vision, bio-acoustics, and neuroscience. An example is shown on the left panel, where you can see a pair of male Zebrafishes (Danio rerio) chilling out, slightly before they start fighting and biting each other. The location and posture of the fishes are tracked and identified, the blue and orange overlaid skeletons, so that we can analyze their behavior with machine-learning algorithms.
I also have a strong interest in evolutionary algorithms and their application in neural architecture search, and more generally artificial intelligence. Have a look around, and if anything comes to your mind, feel free to get in touch through the form at the bottom.
Selected Publications
K. Deligkaris, "Particle Swarm Optimization and Random Search for Convolutional Neural Architecture Search," in IEEE Access, vol. 12, pp. 91229-91241, 2024.
Deligkaris K, Bullmann T and Frey U (2016)
Extracellularly Recorded Somatic and Neuritic Signal Shapes and Classification Algorithms for High-Density Microelectrode Array Electrophysiology. Front. Neurosci. 10:421.
K. V. Deligkaris, Z. D. Zaharis, D. G. Kampitaki, S. K. Goudos, I. T. Rekanos and M. N. Spasos (2009)
Thinned Planar Array Design Using Boolean PSO With Velocity Mutation. IEEE Transactions on Magnetics, vol. 45, no. 3, pp. 1490-1493.