Petia Radeva: «Knowledge transfer is a driving force of motivation for researchers»
The passion for numbers that Petia Radeva had since she was a child led her to obtain a degree in mathematics and computer science, a master’s degree in Image Processing and Artificial Intelligence and a doctorate with a thesis on computer vision. Now, this researcher from Bulgaria based in Barcelona leads the Computer Vision Research Group of the University of Barcelona and is affiliated to the Computer Vision Centre, where she works on the development of algorithms and mathematical and statistical models for the processing of images acquired through lifelogging.
What exactly is ‘lifelogging’?
Lifelogging is the process of capturing images of a person’s life; for example, Facebook is a form of lifelogging because it explains our lives. If we talk about visual lifelogging then we talk about a portable camera, very light and easy to use—with no buttons—that takes two pictures per minute and with which you know everything you’ve done during the day: where you’ve been, who you talked to, what you ate … This paves the way for many applications, because there is increasingly more evidence that lifestyle affects most major diseases. Thanks to this technology, patients can objectively show their doctor if they eat when they say they do, if they go for a walk, if they lead an active life…
What does it have to do with mathematics?
The camera captures two images per minute; therefore, at the end of the day we have two thousand images. Only very few out of these provide information. If we were to revise them every day, at the end of the month we would have up to 100,000 images that no one would watch. We design computational algorithms and mathematical and statistical models to determine whether an image is semantically rich or not. We have developed algorithms that have already been validated and published, and others are in progress.
What applications can this process have?
We are working with the team of Dr. Maite Garolera, at the Terrassa Health Consortium, to implement lifelogging for patients with mild cognitive impairment. These people are still autonomous individuals who can live by themselves at home, but begin to show the first symptoms of a disease that affects their daily lives. Our theory is that using lifelogging and working with their biographical images can exercise their memory and slow down the process. Today these patients solve Sudoku puzzles and mathematical exercises to train their memory, but this is very unnatural for them. We believe that working with biographical images is the most natural exercise, because the emotions behind the images can also help them and make therapies more efficient. Moreover, we also work with issues related to nutrition, migraines, depression, and home assistance for the elderly and for people with disabilities.
Could this technology have some other application outside the health sector?
Lifelogging could be an interesting tool to preserve intangible cultural heritage: customs, trends, traditions, social movements … Portable cameras and social networks gather a lot of information about our cultural life, and processing all this information allows us to save and preserve our intangible culture for future generations.
What do you think is the situation of research in Spain and Catalonia?
I think that the situation is worse than five years ago: the financing of public projects is much lower, less projects are being funded, and their budgets are also lower. There is no money in Spain, so now there is a strong pressure to apply for European projects, but there is no supporting infrastructure. European projects not only demand research technology, but also a business plan and the ability and expertise to handle large groups, which requires strong management capabilities. We will suffer this crisis at least for the next twenty years. Six of the fifteen PhD students I have supervised are working abroad, and the resources that Spain invested in them do no longer benefit the country. The researchers that could lead projects, carry out research, and push technology forward are abroad; we are losing a lot with this situation, but it is difficult to try and convince someone who does not want to listen.
What do you think about knowledge transfer?
I believe strongly in the transfer of knowledge, because, on the one hand, it is a tool for validating all the work we do with theoretical models and, on the other hand, it is a driving force of motivation both for us and for our students. The fact that what we invent, besides being innovative, can help people is very motivating. In addition, there are companies that pose problems which can often be more complex than we could ever imagine. Companies have a direct link with the market, and therefore they can bring a product to market, a task that has little to do with our research.
More about Petia Radeva
The best quality in a scientist: Curiosity
The best invention in history: The computer, because the digitalisation of information has given momentum to our life.
The FBG is…
… very useful for researchers, because it provides the connections with society, institutions, and companies that researchers do not have the time to establish.