Макаров Илья Андреевич
Факультет компьютерных наук
Профессиональные интересы
Должности
- Доцент — Факультет компьютерных наук, Департамент анализа данных и искусственного интеллекта
Био
- · Начал работать в НИУ ВШЭ в 2011 году.
- · Научно-педагогический стаж: 14 лет.
Образование
- 2021 · PhD: Университет Любляны
- 2015 · Аспирантура: Московский государственный университет им. М.В. Ломоносова, факультет: Механико-математический
- 2011 · Специалитет: Московский государственный университет им. М.В. Ломоносова, факультет: Механико-математический, специальность «Математика», квалификация «Математик»
Опыт работы
- · 2011: НИУ ВШЭ, Департамент анализа данных и искусственного интеллекта – старший преподаватель, научный сотрудник ( настоящее время), заместитель руководителя (2012-2017)
Награды и поощрения
- · Номинант на "Лучшие преподаватели 2014" (июль 2014)
- · Надбавка за академическую работу (2017–2018)
- · Надбавка за публикацию в журнале из Списка А (и приравненном к нему научном издании) (2025–2026, 2024–2025, 2023–2024)
- · Надбавка за публикацию в международном рецензируемом научном издании (2022–2023, 2021–2022, 2018–2020)
- · Надбавка за статью в зарубежном рецензируемом научном издании (2016–2017)
- · Лучший преподаватель — 2022, 2017–2018
- · Группа высокого профессионального потенциала (кадровый резерв НИУ ВШЭ)Категория "Новые преподаватели" (2013–2014)Категория "Будущие преподаватели" (2012)
Гранты и проекты
- — · Грант президента РФ МК-5016.2012.1 "Многомерные диофантовы приближения" (2012) - исполнитель
- — · Грант РНФ 17-11-01294 "Представление, обнаружение и обработка знаний: логический подход"
Конференции (10)
Показать все
- · 2016: The 5th international conference on Analysis of Images, Social Networks, and Texts (AIST) (Екатеринбург). Доклад: Smoothing Voronoi-based Path with Minimized Length and Visibility using Composite Bezier Curves
- · 2016: Third International Workshop on Experimental Economics and Machine Learning (EEML 2016) (Москва). Доклад: Modelling Human-like Behavior through Reward-based Approach in a First-Person Shooter Game
- · 2016: The 6th International Conference on Network Analysis (Nizhny Novgorod). Доклад: Co-author Recommender System
- · 2016: ACM Multimedia 2016 (Амстердам). Доклад: First-Person Shooter Game for Virtual Reality Headset with Advanced Multi-Agent Intelligent System
- · 2015: The 4th international conference on Analysis of Images, Social Networks, and Texts (AIST) (Екатеринбург). Доклад: Imitation of human behavior in 3D-shooter game
- · 2015: 10th Panhellenic Logic Symposium (Karlovasi, Samos). Доклад: Total Equivalence Systems for Classes of 3-valued Projection Logic whose Projections Equal to the Class of Linear Boolean Functions
- · 2015: 10th Panhellenic Logic Symposium (Karlovasi, Samos). Доклад: Logical Generalized Continued Fractions
- · 2015: 5th World Congress on Universal Logic (Istanbul). Доклад: Separator Method for Constructing Canonical Types of Formulas
- · 2014: Конференция научно-педагогических работников Национального исследовательского университета «Высшая школа экономики» (Москва). Доклад: Выборы Ученого Совета НИУ ВШЭ
- · 2012: Ломоносовские чтения - 2012 (Москва). Доклад: О некоторых свойствах внутренних полиэдров Клейна
Идентификаторы исследователя
- ORCID:
0000-0002-3308-8825 - ResearcherID:
G-9195-2015 - SPIN РИНЦ:
3151-9176 - Google Scholar: https://scholar.google.com/citations?user=cFpDMzIAAAAJ&hl=en
- Scopus AuthorID:
57203060623
Публикации (117)
Russian Sign Language Dactyl Recognition
2019 · CHAPTER · en
In this paper, we compare several real-time sign language dactyl recognition systems and present a new model based on deep convolutional neural networks. These systems are able to recognize Russian alphabet letters presented as static signs in Russian Sign language used by people from deaf community. In such an approach, we recognize words from Russian natural language presented by consequent hand gestures of each letter. We evaluate our approach on Russian (RSL) sign language, for which we collect our own dataset and evaluate dactyl recognition.
Fast Depth Map Super-Resolution Using Deep Neural Network
2019 · CHAPTER · en
Depth map super-resolution is a challenging computer vision problem. In this paper, we present two deep convolutional neural networks solving the problem of single depth map super-resolution. Both networks learn residual decomposition and trained with specific perceptual loss improving sharpness and perceptive quality of the upsampled depth map. Several experiments on various depth super-resolution benchmark datasets show state-of-art performance in terms of RMSE, SSIM, and PSNR metrics while allowing us to process depth super-resolution in real time with over 25-30 frames per second rate.
Deep Reinforcement Learning in VizDoom First-Person Shooter for Health Gathering Scenario
2019 · CHAPTER · en
In this work, we study the effect of combining existent improvements for Deep Q-Networks (DQN) in Markov Decision Processes (MDP) and Partially Observable MDP (POMDP) settings. Combinations of several heuristics, such as Distributional Learning and Dueling architectures improvements, for MDP are well-studied. We propose a new combination method of simple DQN extensions and develop a new model-free reinforcement learning agent, which works with POMDP and uses well-studied improvements from fully observable MDP. To test our agent we choose the VizDoom environment, which is old first person shooter, and the Health Gathering scenario. We prove that improvements used in MDP setting may be used in POMDP setting as well and our combined agents can converge to better policies. We develop an agent with combination of several improvements showing superior game performance in practice. We compare our agent with Recurrent DQN using Prioritized Experience Replay and Snaphot Ensembling agent and get approximately triple increase in per episode reward.
Deep Reinforcement Learning in Match-3 Game
2019 · CHAPTER · en
An increasing number of algorithms in deep reinforcement learning area creates new challenges for environments, particularly, for their comprehensive analysis and searching application areas. The key purpose of this article is to provide an extensible environment for researches. We consider a Match-3 game, which has simple gameplay, but challenging game design for engaging players. The article provides metrics for evaluation of agents and corresponding baselines in different scenarios.
Predicting Collaborations in Co-authorship Network
2019 · CHAPTER · en
In this paper, we study the problem of predicting collaborations in co-authorship network. We formulated our task in terms of link prediction problem on weighted co-authorship network, in which authors play the role of nodes, and weighted edges connecting two authors are formed by storing either a number or quality metric of research papers co-authored by these authors. Our task is then formulated as regression machine learning model based on network features constructed using network embedding. We evaluate our edge embeddings on large AMiner co-authorship network for (un)weighted node2vec network embeddings and also on the dataset containing temporal information on National Research University Higher School of Economics (HSE) over twenty five years of research articles indexed in Russian Science Citation Index and Scopus for predicting the quality of future research publications measures in terms of quartiles corresponding to published journals indexed in Scopus. We showed that our model of network edge representation has better performance for stated regression task on both, AMiner and HSE co-authorship networks.
On Reproducing Semi-dense Depth Map Reconstruction using Deep Convolutional Neural Networks with Perceptual Loss
2019 · CHAPTER · en
In our recent papers, we proposed a new family of residual convolutional neural networks trained for semi-dense and sparse depth reconstruction without use of RGB channel. The proposed models can be used in low-resolution depth sensors or SLAM methods estimating partial depth with certain distributions. We proposed using perceptual loss for training depth reconstruction in order to better preserve edge structure and reduce over-smoothness of models trained on MSE loss alone. This paper contains reproducibility companion guide on training, running and evaluating suggested methods, while also presenting links on further studies in view of reviewers comments and related problems of depth reconstruction.
Deep Reinforcement Learning with VizDoom First-Person Shooter
2019 · CHAPTER · en
In this work, we study deep reinforcement algorithms for partially observable Markov decision processes (POMDP) combined with Deep Q-Networks. To our knowledge, we are the first to apply standard Markov decision process architectures to POMDP scenarios. We propose an extension of DQN with Dueling Networks and several other model-free policies to training agent using deep reinforcement learning in VizDoom environment, which is replication of Doom first-person shooter. We develop several agents for the following scenarios in VizDoom first-person shooter (FPS): Basic, Defend The Center, Health Gathering. We compare our agent with Recurrent DQN with Prioritized Experience Replay and Snapshot Ensembling agent and get approximately triple increase in per episode reward. It is important to say that POMDP scenario close the gap between human and computer player scenarios thus providing more meaningful justification for Deep RL agent performance.
Efficient Algorithms for Constructing Multiplex Networks Embedding
2019 · CHAPTER · en
Network embedding has become a very promising technique in analysis of complex networks. It is a method to project nodes of a network into a low-dimensional vector space while retaining the structure of the network based on vector similarity. There are many methods of network embedding developed for traditional single layer networks. On the other hand, multilayer networks can provide more information about relationships between nodes. In this paper, we present our random walk based multilayer network embedding and compare it with single layer and multilayer network embeddings. For this purpose, we used several classic datasets usually used in network embedding experiments and also collected our own dataset of papers and authors indexed in Scopus.
Deception Detection in Online Media
2019 · CHAPTER · en
Russian Federation and European Union are fighting against fake news together with other countries in various topics. The disinformation affected British referendum of existing EU, the US election and Catalonia’s referendum are broadly studied. A need for automated factchecking increases, European Commission’s Action Plan 8 is an evidence. In this work, we develop a model for detecting disinformation in Russian language in online media. We use reliable and unreliable sources to compare named entities and verbs extracted using DeepPavlov library. Our method shows four time greater recall compared to chosen baseline.
Deep Reinforcement Learning Methods in Match-3 Game
2019 · CHAPTER · en
A large number of methods are being developed in the deep reinforcement learning area recently, but the scope of their application is limited. The number of environments does not always allow for a comprehensive assessment of a new agent training algorithm. The main purpose of this article is to present another environment for Match-3 game that could be expanded, which would have a connection with the real business. The results for the most popular deep reinforcement learning algorithms are presented as a baseline.
Курсы (7)
-
Research Seminar in Financial Economics
2025/2026 · Магистратура · Анг
-
Литература Древнего Египта
2024/2025 · Бакалавриат · рус
-
Visual geometry and 3D image processing
2022/2023 · Маго-лего / Нижний Новгород · Анг
-
Network Science
2021/2022 · Магистратура · Анг
-
Project Seminar ''Intelligent Systems and Structural Analysis''
2021/2022 · Магистратура · Анг
-
Social Networks
2021/2022 · Магистратура · Анг
-
Structural Analysis and Visualization of Networks
2021/2022 · Магистратура · Анг