Search by tag «Neural Networks» 62 results

• Design in the Era of AI: Skills to Learn to Stay Relevant on the Job Market

  According to the recent Future of Jobs study by the World Economic Forum, graphic design will be among the ten fastest-declining job niches in the coming years. No wonder: neural networks do quite a good job on illustrations, banners, and logos. What should human graphic designers do? What skills can they gain to retain their jobs? And can AI fully replace a human specialist? For answers, we turned to Sergey Gurov, a specialist with 20 years of experience in design, a lecturer at Bang Bang Education, and a co-founder of the OSNOVA workshops.

  20.10.2025

• ITMO Researchers Develop Affordable, Versatile Model for Tracking Potential Production Accidents

  A team of scientists from ITMO University has developed software that can automatically track potential human causes of industrial accidents, such as drinking or smoking in the workplace. Unlike its alternatives, the model is capable of detecting up to 10 events simultaneously, has the accuracy of 80% (higher than that of VideoMAE, the best open-source model trained on the same data), and requires less computational resources. The algorithm is already in use at a major production site in Perm Krai, where it has lowered the required number of in-person checkups by threefold.

  08.09.2025

• ITMO PhD Students Train Algorithms To Count Cells in Microalgae Images With Computer Vision

  Researchers from ITMO University have developed a new method for automated assessment of microalgae cell concentration in a solution based on microscope images. Unlike other popular neural networks, this method uses classic computer vision algorithms and doesn’t require training data. The new solution can be launched on a standard PC and can determine cell concentration in a sample in 30 seconds instead of the usual 30 minutes, with accuracy equal to manual calculation. This algorithm will facilitate environmental studies and improve their accuracy.

  11.08.2025

• Optical Metamaterials Can Become More Accessible – With Matrix Shrinker

  ITMO researchers discovered a way to simplify and accelerate the complex calculations required to generate metasurfaces and periodic optical structures. The mathematical tool called Tensor Train will enable quick development of optical computers, neural networks, and other devices that need precise calculations of light properties. The findings of the study funded by the Russian Science Foundation were published in Computer Physics Communications.

  17.02.2025

• AI Levels Up: What Makes DeepSeek So Powerful

  The Chinese startup DeepSeek has recently unveiled its eponymous AI model. In just a couple of days, it managed to plummet the stocks of big tech companies, landing at the top of most-downloaded apps throughout the world. What can DeepSeek do, how is it superior to ChatGPT, and what new avenues does it open to Russian developers? In this overview, we discuss these questions with scientists from ITMO’s Institute of Applied Computer Science.

  29.01.2025

• Nobel Prize in Physics Awarded for Discoveries in Machine Learning and Neural Networks

  This year, the Nobel Prize in Physics was awarded to John J. Hopfield and Geoffrey E. Hinton “for foundational discoveries and inventions that enable machine learning with artificial neural networks.” In this article, Anton Kuznetsov, the head of ITMO’s Institute of Applied Computer Science, explains the significance of their work – and how it led to ChatGPT.

  08.10.2024

• AI for the Classroom: Tips and Tools

  AI is not just a toy for generating images. According to Gartner, by 2025 around 80% of IT products will be developed with AI. Businesses use bots for client communication, launch neural network-based apps, and deploy algorithms to test cybersecurity systems. Educators, too, can benefit from the new technologies. How? Read on to find out with tips from ITMO lecturers Elena Ponomareva, Artem Sigaev, and Oksana Nakonechnaya.

  04.09.2024

• ITMO Researchers Discover Crystals That Imitate Brain Cell Activity

  Scientists from ITMO University have discovered that when subjected to laser pulses, hybrid crystals behave like neurons in the brain: their reaction to light is analogous to that of neurons to triggers. Such crystals can serve as a basis for devices that work as biological neural networks – thus making it possible to train artificial neural networks more quickly and efficiently. The suggested crystals are made from harmless domestically made materials and are easy to manufacture. The results of the study were described in Communication Materials (part of the Nature journal family).

  14.08.2024

• Tiny Orchestra: New AI Model Uses Water Vapor to Make Music

  In collaboration with researchers from Tyumen State University, scientists at ITMO have created an algorithm that can decipher the behavior patterns of droplets inside a cloud dispersed over water. Research into these processes opens opportunities to create computers that can program living systems, such as bacteria clusters, or solve complex tasks unavailable to conventional computers. What’s more, the suggested neural network can transform the data into music, making it possible to “hear” the particles interacting. The resulting publication describes the first observation of particle reactions inside a levitating droplet cluster.

  01.08.2024

• AI Teaches Sberbank Employees To Speak in Public

  A team of developers helmed by ITMO graduate Yulia Khlyupina has created a virtual coach that offers personalized public speaking courses. The team received a grant of 1 million rubles from Sberbank and 12 million rubles in funding at the Demo Day of the SberStudent accelerator for students and university staff. The robot is currently being trained to offer test services for the company’s employees.

  12.02.2024