Students presented the best IT ideas at this year's Excel@Fit conference hosted online

On Wednesday, the sixth year of the annual Excel@FIT student conference focusing on innovations, technology and research in the IT field was held at the Faculty of Information Technology. The conference presents the best ideas of the students of information technology. Due to the extraordinary situation, the presentations of students' works were held online.

An expert panel comprising professionals from both the academic and commercial sectors chose over 30 works for the final presentations at the Excel@FIT 2020 conference. The aim of the conference is to show practical solutions developed by students of IT; examples of the works presented by the students this year include an application that helps its users to pick a colour before painting their room, a tool detecting errors in programmes, and a platform that is used for exposing criminals using cryptocurrencies for their illegal activities.

Dominik Vagala presented his Android application that enables its users to try out different wall colours in real time. "I wanted to paint my room about two years ago and I just couldn't pick the right colour. It is not that easy to imagine which colour would be best suited for one's room. I tried existing applications, but they were mostly incompatible with my mobile device. That is why I had to use a graphic editor on my computer to change the colours there. This process took too much time and I especially lacked the ability to walk around the room and look at walls with different colours from various angles," Vagala describes his motivation. He managed to finish his application, named 'Paint my Room', and publish it on the Google Play store.

Vladislav Bambuch created a functional platform that can help expose criminals using cryptocurrencies for their illegal activities. His work is using data mining from publicly available websites. "The data-mining process can take months so I primarily focused on its acceleration. I came by this topic during a work on a university project. I found this issue, which could help increase the security on the Internet, quite fascinating and saw an opportunity to build something with a practical use as part of my diploma thesis project," Bambuch says. The result of his work is a functional platform that serves as a basis for the future connection of other web-based tools and their integration with other projects undertaken at FIT BUT.

Roman Bártl designed a device capable of displaying various notifications from telephones similarly to what smart watches do. Incoming calls, text messages or notifications from other applications are shown as icons on a matrix display built from LEDs. "I have created a functional prototype so far. The work also included development of an application that manages the pairing of the phone with the device and where the user can set which notifications he/she wants to be displayed and what icon should represent them," Bártl describes.

Michal Ormoš co-operates with Brno-based company, Sewio on the development of a system enabling localisation of people in buildings. "We are all used to using our phones to help us with navigation, we can use GPS to 'find' each other anywhere in the open space and choose the right direction. However, this does not work once we enter a building because our phones need an unobstructed connection with the satellites in the sky," he notes. Technology for localisation inside buildings, based on ultra-wideband radio transmission using specialised hardware, is gradually becoming embedded in the newest smart phones and, unlike GPS, it enables to localise devices even in places where it was not possible before. "With this technology, for example, during the current pandemic, building managers would be able to monitor the movement of people in a building, check that they observe proper social distancing and, in case of confirmed infection, find out who the infected person was in contact with in the past 24 hours," Michal Ormoš explains. His year-long work on the project yielded a test kit enabling automated validation of this technology across its development.

Monika Mužikovská was also one of the presenters at the conference. She managed to expand the ANaConDA framework to enable detection of errors in concurrent programmes. "These errors may be very serious and actually caused deaths in the past - for example, in the infamous case of Therac-25 radiation therapy machine, an error in parallel programming caused massive overdose of radiation in several patients. ANaConDA and other similar tools can detect errors in parallel programmes using fibres to achieve concurrency. However, there is another large group of multi-process programmes that the existing tools cannot analyse," Mužikovská explains.

FIT students were not the only ones to present their work at the conference. Lukáš Dobiš, a student of biomedical engineering at FEEC BUT, presented his fast, robust and modular solution for detection of persons and recognition of their features on the basis of visual data. "First, we detect the position of the person's face and then we can determine the sex, emotional state or age of the person. We used a combination of several convolutional neural networks for the detection of faces and analysis of their features. A single network was used for each sub-task," says Dobiš. He adds that he had always been interested in artificial intelligence and as the impact of neural networks on our everyday lives increases, he wanted to analyse the technology and find its limits.

Panel of experts awarded prizes to a total of 20 students' works. The Jiří Kunovský Award was awarded to six projects based on over 900 votes cast by public in an online voting. Another twenty prizes were awarded by the industrial partners. The award-winning authors received stipends in the total value exceeding 180 thousand Czech crowns. "The Excel@FIT conference enables students to present their ideas and results of their work. Unfortunately, the extraordinary measures adopted in response to the current situation prevented us from hosting the conference as we were used to in the past. We had to cancel some parts of the programme altogether and other parts were moved to the digital world. Despite the various restrictions, majority of the students worked really hard on their projects and we are very happy with the extent and quality of our students' results," notes Vítězslav Beran, Vice-dean for External Relations at FIT BUT and Executive Chairman of the Excel@FIT conference board.

You can go HERE to see the results as well as excerpts from the students' works.