MediaEval is a benchmarking initiative establishedin2008. Within the Mediaeval Benchmark, several sub benchmarks exist. One of them is the Medico task which is especially focused on medical multimedia systems. In the 2019 edition of the task, the goal was to perform prediction from sperm videos and patient-related data.
Jon Magnus, a first-year bachelor studentin mathematics at the University in Oslo, participated in the task.His solution to the task was based on deep neural networks. For his solution, he received thedistinctive mentioning of Youngest participantwith a great idea on how to solve the task.
At SimulaMet, Jon Magnus is mainly working on Machine Learning for automatic human spermatozoa classification. In specific he is developing deep neural networks for analyzing videos from the sperm samples
To read more about the MediaEval initiative click here.