Researchers from HSE University in Perm Teach AI to Analyse Figure Skating
Researchers from HSE University in Perm have developed NeuroSkate, a neural network that identifies the movements of skaters on video and determines the correctness of the elements performed. The algorithm has already demonstrated success with the basic elements, and further development of the model will improve its accuracy in identifying complex jumps.
Figure skating is a technically challenging sport in which not only speed and strength are important, but so is precision of movement. Coaches rely on their eyes and experience, but objective data can provide more information on which elements work better, which require improvement, and how technique changes over time. This is especially valuable in children's sports, where groups can have as many as 15 people and it is difficult for coaches to monitor each athlete at the same time.
The team of the NeuroSkate project, which includes Anna Provorova, Daria Semenova, Lyudmila Gergert, Sofya Kulikova, and Irina Polyakova, decided to test how artificial intelligence would cope with analysing the movements of figure skaters. The researchers selected six movements that are easy to identify by an athlete's posture: Biellmann, spin, boat, and several basic single jumps—flip, Rittberger, and Lutz. The main idea was that the neural network could automatically find the skater in the video, track their movements, and mark the performed element.
‘It sounds simple, but in practice it’s not an easy task. Large databases are required to train algorithms, and very few of them are open source. Research analyses adult athletes, but no one has collected videos of young figure skaters with marked movements so far,’ says Anna Provorova, Junior Research Fellow at the HSE Centre for Cognitive Neuroscience in Perm.
The project was implemented in collaboration with the Orlyonok Sports School of the Olympic Reserve in Perm, which provided data for training the neural network.
‘While the skaters were on their summer break, we used competition recordings and publicly available videos. Later, we managed to take our own training footage, but only on a phone, without professional equipment, which affected the image quality and marking accuracy,’ says Sofya Kulikova, Director of the HSE Centre for Cognitive Neuroscience in Perm.
Motion recognition includes several stages. First, the video is divided into frames, each of which highlights key points on the skater's body. Then a sequence of 60 frames with marked dots is transmitted to a neural network that analyses the athlete's movements.
First, the model was tested in binary classification: it was offered pairs of movements to distinguish, such as a Rittberger and a boat, without adding other elements. In this format, the system worked stably, showing an accuracy of 72%. But as soon as the task became more complicated and more elements were added, the system started to make mistakes. The researchers continued to mark up new videos and further train the model, making it easier to recognise athletes in the video.
The developers have also created a web application that allows users to upload training videos and analyse the statistics of a particular athlete. This could become a tool that helps coaches monitor student progress without having to review hours of recordings.
‘We hope to keep working on the project, as we have ideas on how to improve the algorithm: the most important thing is to collect a large and high-quality database of athletes' videos. There is also an understanding of how to improve the motion recognition process, for example, using graph neural networks. This is a very promising area. One of these models (HD-GCN) showed impressive results, which were presented at the International Conference on Computer Vision in 2023. However, it has not yet been possible to launch it on real project data,’ says Anna Provorova.
The study was implemented as part of the Priority 2030 programme.
See also:
Neurolinguists Assist in Awake Surgery on 11-Year-Old Patient with Epilepsy
Researchers at the HSE Centre for Language and Brain took part in a rare awake neurosurgical procedure performed on an 11-year-old patient with drug-resistant epilepsy. Working alongside surgeons at the Voyno-Yasenetsky Centre of Specialised Medical Care for Children in Solntsevo, they monitored the resection of a portion of the left temporal lobe, where the epileptic focus had been identified.
Scientists Explain How Emotions Shape Attitudes Toward Digital Governance
Today, interactions between citizens and government increasingly take place through digital governance platforms, including digital public services, AI-powered systems, and algorithmic decision-making tools. Until now, however, these technologies have largely been viewed as technical instruments, with their effectiveness assessed primarily in terms of efficiency and user-friendliness. The authors of a new study propose a broader perspective, arguing that digital governance should also be understood as an emotional experience that directly shapes citizens' trust in public institutions.
Neural Network Maps as a Method for Constructing Mathematical Models
Scientists from HSE University–Nizhny Novgorod and the Institute of Physics Belgrade, Serbia, are jointly exploring the application of machine learning techniques and neural networks to the study of nonlinear dynamics. Natalya Stankevich, Leading Research Fellow at the Laboratory of Topological Methods in Dynamics of the Faculty of Informatics, Mathematics, and Computer Science at HSE University–Nizhny Novgorod, spoke to the HSE News Service about this international project.
HSE Scientists Develop Method to Compress Large Language Models Without Losing Quality
Researchers from the AI and Digital Science Institute at the HSE Faculty of Computer Science have developed a new compression method for large language models such as GPT and LLaMA that reduces their size by 25–36% without additional training or significant loss of accuracy. This is the first approach to use mathematical transformations—specifically, rotations of model weights—to make models more amenable to compression with structured matrices. The study results have been published in ACL Findings 2025. The code is available on GitHub.
Machine Learning Models Can Help Reduce Volatility and Boost Stock Market Returns
The use of machine learning models makes it possible to achieve greater accuracy in predicting risks in the Russian stock market compared to classical econometric approaches. The predictive power of these models increases by 23%, while the average investor’s return can reach up to 13% per annum. These conclusions were drawn by Nikita Lysenok from the Department of Financial Market Infrastructure at the HSE Faculty of Economic Sciences. The paper has been published in Fundamental and Applied Mathematics.
Pocket Money, Personal Interest, and Family Practices: What Shapes Students’ Economic Literacy?
University students' economic literacy depends not only on their field of study but also on their interest in economics, the learning environment, and family financial practices. For example, students who received pocket money irregularly tend to perform better on economic literacy tests than their peers who received financial support on a regular basis. These findings come from a study conducted by HSE University involving more than 1,100 students from five Russian universities. The findings have been published in Cakrawala Pendidikan.
HSE Study Reveals Imbalance in the Generative AI Market
Researchers at HSE University analysed how effectively the global generative artificial intelligence market converts investment into real revenue, concluding that AI is currently developing faster than it is paying off. The results have been published in the journal Foresight and STI Governance.
HSE Scientists Train Neural Network to 'Hear' Faults in Electric Motors
Researchers at the AI and Digital Science Institute of the HSE Faculty of Computer Science have developed a new method—the Signature-Guided Data Augmentation (SGDA) framework—that achieves 99% accuracy in motor fault detection and 86% accuracy in fault classification. The application of this approach can reduce industrial equipment repair costs, minimise downtime, and improve production safety. The study results have been published in Engineering Applications of Artificial Intelligence.
The 'Second Shift' Is Not Why Women Avoid News
Women are more likely than men to avoid political and economic news, but the reasons for this behaviour are linked less to structural inequality or family-related stress than to personal attitudes and the emotional perception of news content. This conclusion was reached by HSE researchers after analysing data from a large-scale survey of more than 10,000 residents across 61 regions of Russia. The study findings have been published in Woman in Russian Society.
Resource Race and Green Transition: Three Unexpected Conclusions from Foresight Centre’s Research on Climate and Poverty
Beneath the surface of green energy—which most people associate with solar panels, electric vehicles, and reduced CO2 emissions—lies a complex web of geopolitical interests, international inequality, and resource constraints. Researchers from the Laboratory for Science and Technology Studies (LST) at the HSE ISSEK Foresight Centre have published a series of articles in leading international journals on hidden and overt conflicts surrounding critically important metals and minerals, as well as related processes in the energy sector.