Reading human emotions through eye movements

Recognizing human emotions and identifying people through eye movement tracking is a project that scientists from the Institute of Information and Communication Technologies at the Bulgarian Academy of Sciences work on. The scientific team includes young researcher Teodor Vakarelski, who was inspired by the idea five years ago, attending a lecture on Eye Tracking by his scientific supervisor, Prof. Kiril Andreev.

"What information can we extract from eye movements - that was the initial question we asked ourselves,” Teodor Vakarelski recalls. "In our laboratory, we put a so-called eye tracker device on people. This is a high-frequency 1000 Hz camera. The camera takes a thousand pictures per second with very good resolution and works in the infra-red spectrum. Thanks to its extremely high accuracy, we can record and analyse even the smallest eye movements.”

During the experiments, the volunteers had to focus on a certain point. And although it seems that their eyes do not twitch, it turns out that they perform three types of movements, one of which is suitable for analysis for identification purposes. According to Teodor Vakarelski, the method guarantees much greater security, since a fingerprint and signature could be forged.

The scientists' activities are part of the European VIBraTE project of Assoc. Prof. Dimitar Prodanov with the main goal of developing and improving deep brain implants.

"In this way, a brain-computer system is created, which examines the interaction between machine and human,” Teodor Vakarelski says in an interview for the "Hristo Botev" program of the Bulgarian National Radio. “One of the sub-directions of the project is non-invasive systems - they do not cause limitations to the person and it is not necessary to use a brain implant to obtain the information. This can be done with a remote device that scans it."

For this purpose, the institute has already built the first laboratory in Bulgaria for testing brain implants, equipped with an optical table that is well isolated from environmental vibrations. There, one can simulate pulse, breathing, various movements and study the connection between the human brain and the brain implant. "Our goal is to create a model with better characteristics compared to the current state of brain implants, as they hide many dangers and imperfections," the scientist adds.

The interdisciplinary research team hopes to have a final result by the end of next year – a model that will improve the current state of the technology and be usable.

"Many diseases could be positively affected by brain implants," says Teodor Vakarelski. "Some of the models are related to diagnostics, others to improving the condition, but their main application at this stage is in paralysis of the body - when the brain functions well, but as a result of an accident, a nerve in the spine is severed, leaving the person paralyzed. In fact, the brain implant allows information from the brain to be taken, transmitted to artificial intelligence for analysis, after which a cursor can be moved on a computer screen and even an exoskeleton or wheelchair can be set in motion, i.e. the person can be mobile."

According to the scientist, with the help of a brain implant, it is possible to predict when there is a risk of an epileptic shock - in such a situation we should stop driving and lie down. Clinical depression is also affected by electro stimulation, without causing addiction as is the case with most medications.

Text: Diana Tsankova (based on an interview with Ani Kostova from the "Hristo Botev" program of the BNR)