Researchers from HSE University and Artificial Intelligence Research Institute Train Neural Network to Learn Much More Efficiently
Neural network generative models have achieved impressive results in recent years, but researchers are still working on increasing their efficiency. Researchers from the HSE Faculty of Computer Science and AIRI have managed to optimize the finetuning of the StyleGAN2 neural network, which creates realistic images, to new domains, reducing the number of trained parameters by four orders of magnitude. At the same time, the quality of the obtained images remained high. The results of this work were presented at the NeurIPS 2022 conference.
Pet a Robocat and Talk to R2D2: HSE Faculty of Computer Science Holds Robotics Fest
At the end of April, the HSE Faculty of Computer Science organised its first robotics festival. SBER was a partner of the event, which took place in the atrium of the HSE building on Pokrovsky Bulvar. Yandex and the National Association of Robotics Market Participants were the event's co-organisers. The guests attended lectures from industry experts and interacted with robots on interactive platforms.
HSE University-VK Engineering and Mathematics School Launches Master's Programme in AI Technologies
The Master’s Programme ‘Applied Artificial Intelligence Models’ trains professionals with the competencies to create industrial artificial intelligence models capable of solving a wide range of applied tasks. During training, students will study and apply modern methods and means of creating applied models using artificial intelligence technologies.
Artificial Case: ChatGPT Helps HSE Lawyer Write Article for Academic Journal
Roman Yankovskiy, head of the Centre for Legal Education Transformation at the HSE Faculty of Law, academic supervisor of the Digital Lawyer programme, has written an article in a legal journal using ChatGPT. The experiment was agreed upon with the journal’s editorial board. The author explains why we need such an experience, whether artificial intelligence (AI) is up to the task, and whether we can prohibit its use in scientific work.
‘AI Is a Tool, and Those Who Master It Will Have a Competitive Advantage’
HSE University is hosting the FIT-M 2022 International Scientific Forum. As part of this event, the HSE Cultural Centre held a series of lectures on November 29–30. On December 7–9, guests of the forum will have three days of practical work with scientists, IT industry leaders, businesspeople, and industry experts.
HSE Researchers Contribute to Artificial Intelligence Journey Conference
The AI Journey international conference is a major platform for sharing cutting-edge innovations in artificial intelligence and machine learning. In late November 2022, AIJ was once again hosted by Sber. The conference was attended by HSE researchers from the Faculty of Computer Science and the Centre for Artificial Intelligence.
Fall into ML: Autumn School and Conference on Machine Learning Held at HSE University
On November 1st-3rd, 2022 the International Laboratory of Stochastic Algorithms and High-Dimensional Inference of the HSE Faculty of Computer Science and the Laboratory of Methods for Big Data Analysis with the support of HSE AI Centre and the Russian Science Foundation organized the first autumn school and conference on artificial intelligence ‘Fall into ML’. The new format of the event included a school for students and young researchers.
Russian Scientists Teach AI to Analyse Emotions of Participants at Online Events
HSE researchers have proposed a new neural network method for recognising emotions and people's engagement. The algorithms are based on the analysis of video images of faces and significantly outperform existing single models. The developed models are suitable for low-performance equipment, including mobile devices. The results can be implemented into video conferencing tools and online learning systems to analyse the engagement and emotions of participants. The results of the study were published in IEEE Transactions on Affective Computing.
AI Worse at Recognizing Images than Humans
Researchers from HSE University and Moscow Polytechnic University have discovered that AI models are unable to represent features of human vision due to a lack of tight coupling with the respective physiology, so they are worse at recognizing images. The results of the study were published in the Proceedings of the Seventh International Congress on Information and Communication Technology.
Scientists Doubt that DeepMind’s AI Is as Good for Fractional-Charge Systems as it Seems
In their paper published in Science in 2021, a DeepMind team showed how neural networks can be used to describe electron interactions in chemical systems more accurately than existing methods. A team of researchers from Skoltech, the Zelinsky Institute of Organic Chemistry, HSE University, Yandex, and Kyungpook National University show in their comment in Science that DeepMind AI’s ability to generalise the behaviour of such systems does not follow from the published results and requires revisiting, the Skoltech website says.