|
ИСТИНА |
Войти в систему Регистрация |
ИСТИНА ПсковГУ |
||
Perturbation-Based Analysis of the R(2+1)D Network and a Deformable Extension for Radar-Frequency Image Classificationстатья
Статья опубликована в журнале из списка RSCI Web of Science
Статья опубликована в журнале из перечня ВАКСтатья опубликована в журнале из списка Web of Science и/или Scopus
Дата последнего поиска статьи во внешних источниках: 15 апреля 2026 г.
- Авторы: Sargsyan V., Mkrtchyan G., Saroyan E., Aslanyan L.
- Журнал: Pattern Recognition and Image Analysis: Advances in Mathematical Theory and Applications
- Том: 36
- Номер: 1
- Год издания: 2026
- Издательство: Pleiades Publishing, Ltd
- Местоположение издательства: Road Town, United Kingdom
- Первая страница: 90
- Последняя страница: 109
- DOI: 10.1134/S1054661825701330
- Аннотация: Target identification is one of the fundamental challenges in radar systems, involving the classification of detected objects based on their type (e.g., human, vehicle, etc) and motion state (stationary or moving). Traditional approaches rely on analyzing variations in the target's motion vector parameters. However, advancements in computational power and parallel processing algorithms have enabled the integration of artificial intelligence (AI) methods to enhance identification accuracy and efficiency. While AI-based approaches require extensive and computationally intensive training phases, their inference process is typically fast and scalable.This paper presents a deep learning approach for the classification of radio-frequency (RF) images using a Residual (2+1)D Convolutional Neural Network (CNN), further enhanced with a Long Short-Term Memory (LSTM) layer to model temporal dependencies. To interpret and evaluate the model’s decision-making process, we employ the Spatio-Temporal Extremal Perturbation (STEP) method, which reveals key regions influencing classification outcomes.
- Добавил в систему: Саргсян Ваге Гнелович