Search by tag «Neural Networks» 42 results
Is it true that neural networks can’t tell a gorilla from a tumor? What is prediction confidence and what do they have to do to find your car in the parking lot? And finally, why shouldn’t you trust a model’s conclusions completely? We addressed all of these questions to Natalia Khanzhina, a PhD student at ITMO’s Information Technologies and Programming Faculty, who developed a new method that improves the precision of the neural network that detects brain tumors in MRI scans.
First envisioned in the 1950s, neural networks are now all the rage: they’ve taken over the media and even your day-to-day life. After having a few profound conversations with Siri or Alexa, enjoying some perfectly-picked tunes on your personalized radio, or unexpectedly seeing your long-lost friend on your social media feed, you may start to wonder what else is in the power of neural networks.
This year’s Congress of Young Scientists brought together 114 school students from 33 cities in Russia and Belarus. Most of the participants will graduate this year and many took part in a conference of such magnitude and level for the first time. As a result, experts chose 15 winners who will receive prizes and advance to the final round of the ITMO.STARS contest, as well as three more holders of the special award.
Computers can outperform people in many fields – but, it seemed, not in the art of making jokes. ITMO University researchers decided to try and dispute this claim. This year, to celebrate April Fools’ Day they taught GPT-3, the world’s most powerful text-generating algorithm, to make jokes. Now anyone on the internet can test the results of their work.
A team of scientists from ITMO University has created a model for an algorithm that can analyze the meaning of texts. It differs from regular neural networks because it not only takes into account word usage frequency but also introduces a category of subjective text perception. Read on to find out what it means for the future of artificial intelligence and how quantum theory is of help here.
Prof. Aleks Farseev and his colleagues will get access to the GPT-3 neural network that can produce content that blurs the lines between machine and human crafted content. The team will use the new capabilities for improving advertising algorithms in social media.
Having secured the special Seed Grants funding from the university’s Digital Humanities Research Center, ITMO students and staff members are launching a new popular science project with a focus on history. There, they teach neural networks to “reimagine” the letters sent from the battlegrounds during World War II to attract the public to the challenge of keeping alive the memory of the WWII heroes and victims.
ITMO researchers together with a colleague from the University of Oulu in Finland created an algorithm for predicting the success of new books. The algorithm's claimed efficacy is based on the analysis of the emotional fluctuations of a literary text.
Every day, billions of people use social media to communicate with each other. But some of them doxx others, insult them, and act in a toxic manner. ITMO.NEWS met with Ivan Smetannikov, the co-head of ITMO’s Machine Learning Lab, to find out how algorithms help us combat this behavior.
ITMO Scientists Use Computer Vision Technologies and Neural Networks to Assess Meditation Efficiency
The algorithm created by ITMO researchers can assess a person’s psycho-emotional state based on a smartphone video of a meditation session. It tracks evenness of breath, body positioning and limb movements. The research results were published in Future Internet.