Макаров Илья Андреевич

Prediction of News Popularity via Keywords Extraction and Trends Tracking

2021 · CHAPTER · en

In the last years, news agencies have become more influential in various social groups. At the same time, the media industry starts to monetize online distributed articles with contextual advertising. However, the efficiency of online marketing highly depends on the popularity of news articles. In our work, we present an alternative and effective way for article popularity forecasting with two–step approach: article keywords extraction and keywords-based article popularity prediction. We show the benefits of this technique and compare with widely used methods, such as Text Embeddings and BERT–based methods. Moreover, the work provides an architecture of the model for dynamic keyword tracking trained on the newest dataset of Russian news articles with more than 280k articles and 22k keywords for the popularity of forecasting purposes.

Federated Learning in Named Entity Recognition

2021 · CHAPTER · en

This article is devoted to the implementation of the federated approach to named entity recognition. The novel federated approach is designed to solve data privacy issues. The classic BiLSTM-CNNs-CRF and its modifications trained on a single machine are taken as baseline. Federated training is conducted for them. Influence of use of pretrained embedding, use of various blocks of architecture on training and quality of final model is considered. Besides, other important questions arising in practice are considered and solved, for example, creation of distributed private dictionaries, selection of base model for federated learning.

Fusion of text and graph information for machine learning problems on networks

2021 · ARTICLE · en

Today, increased attention is drawn towards network representation learning, a technique that maps nodes of a network into vectors of a low-dimensional embedding space. A network embedding constructed this way aims to preserve nodes similarity and other specific network properties. Embedding vectors can later be used for downstream machine learning problems, such as node classification, link prediction and network visualization. Naturally, some networks have text information associated with them. For instance, in a citation network, each node is a scientific paper associated with its abstract or title; in a social network, all users may be viewed as nodes of a network and posts of each user as textual attributes. In this work, we explore how combining existing methods of text and network embeddings can increase accuracy for downstream tasks and propose modifications to popular architectures to better capture textual information in network embedding and fusion frameworks.

Semi-automatic Manga Colorization Using Conditional Adversarial Networks

2021 · CHAPTER · en

Manga colorization is time-consuming and hard to automate. In this paper, we propose a conditional adversarial deep learning approach for semi-automatic manga images colorization. The system directly maps a tuple of grayscale manga page image and sparse color hint constructed by the user to an output colorization. High-quality colorization can be obtained in a fully automated way, and color hints allow users to revise the colorization of every panel independently. We collect a dataset of manually colorized and grayscale manga images for training and evaluation. To perform supervised learning, we construct synthesized monochrome images from colorized. Furthermore, we suggest a few steps to reduce the domain gap between synthetic and real data. Their influence is evaluated both quantitatively and qualitatively. Our method can achieve even better results by fine-tuning with a small number of grayscale manga images of a new style. The code is available at github.com.

Automated Image and Video Quality Assessment for Computational Video Editing

2021 · CHAPTER · en

We study non-reference image and video quality assessment methods, which are of great importance for computational video editing. The object of our work is image quality assessment (IQA) applicable for fast and robust frame-by-frame multipurpose video quality assessment (VQA) for short videos. We present a complex framework for assessing the quality of images and videos. The scoring process consists of several parallel steps of metric collection with final score aggregation step. Most of the individual scoring models are based on deep convolutional neural networks (CNN). The framework can be flexibly extended or reduced by adding or removing these steps. Using Deep CNN-Based Blind Image Quality Predictor (DIQA) as a baseline for IQA, we proposed improvements based on two patching strategies, such as uniform patching and object-based patching, and add intelligent pre-training step with distortion classification. We evaluated our model on three IQA benchmark image datasets (LIVE, TID2008, and TID2013) and manually collected short YouTube videos. We also consider interesting for automated video editing metrics used for video scoring based on the scale of a scene, face presence in frame and compliance of the shot transitions with the shooting rules. The results of this work are applicable to the development of intelligent video and image processing systems.

Community Detection Based on the Nodes Role in a Network: The Telegram Platform Case

2021 · CHAPTER · en

The paper studies the community detection problem on Telegram channels. The dataset is received from TGStat service and includes the information of 58k forwards between 100 politician Telegram channels. We implement modern clustering approaches to solve the problem of missing social links. Our study is based on a combination of structural features with strategy-based attributes, including indicators designed according to the nodes’ role in a network. Authors provide ten novel indicators, which are calculated for each network’s member per each message in order to vectorize a Telegram channel with regard to its strategy of information spread and the way of contacting other channels. Authors construct a metric-based graph of channel relations and cluster channels representations using network science techniques. Obtained results are studied using quantitative and qualitative analysis showing promising results in applying joint network-based and KPI-based models for the stated problem.

Human Action Recognition for Boxing Training Simulator

2021 · CHAPTER · en

Computer vision technologies are widely used in sports to control the quality of training. However, there are only a few approaches to recognizing the punches of a person engaged in boxing training. All existing approaches have used manual feature selection and trained on insufficient datasets. We introduce a new approach for recognizing actions in an untrimmed video based on three stages: removing frames without actions, action localization and action classification. Furthermore, we collected a sufficient dataset that contains five classes in total represented by more than 1000 punches in total. On each stage, we compared existing approaches and found the optimal model that allowed us to recognize actions in untrimmed videos with an accuracy 87%.

Deep Reinforcement Learning in VizDoom via DQN and Actor-Critic Agents

2021 · CHAPTER · en

In this work, we study the problem of learning reinforcement learning-based agents in a first-person shooter environment VizDoom. We compare several well-known architectures, such as DQN, DDQN, A3C, and Curiosity-driven model, while highlighting the main differences in learned policies of agents trained via these models.

Hotel Recognition via Latent Image Embeddings

2021 · CHAPTER · en

Fast Depth Reconstruction Using Deep Convolutional Neural Networks

2021 · CHAPTER · en

In this paper, we study depth reconstruction via RGB-based, Sparse-Depth, and RGBd approaches. We showed that combination of RGB and Sparse Depth approach in RGBd scenario provides the best results. We also proved that the models performance can be further tuned via proper selection of architecture blocks and number of depth points guiding RGB-to-depth reconstruction. We also provide real-time architecture for depth estimation that is on par with state-of-the-art real-time depth reconstruction methods.