Перегляд за Автор "Smilianets, F."

Зараз показуємо 1 - 3 з 3
  • Ескіз
    ДокументВідкритий доступ
    Application of embeddings for multi-class classification with optional extendability
    (КПІ ім. Ігоря Сікорського, 2024) Smilianets, F.
    This study investigates the feasibility of an expandable image classification method, utilizing a convolutional neural network to generate embeddings for use with simpler machine learning algorithms. The possibility of utilizing this approach to add new classes by additional training without modifying the topology of the vectorization network was shown on two datasets: MNIST and Fashion-MNIST. The findings indicate that this approach can reduce retraining time and complexity, particularly for more complex image classification tasks, and also offers additional capabilities such as similarity search in vector databases. However, for simpler tasks, conventional classification networks remain more time-efficient.
  • Ескіз
    ДокументВідкритий доступ
    Application of transfer learning for enhanced pulmonary disease detection via ct image embeddings
    (КПІ ім. Ігоря Сікорського, 2024) Smilianets, F.
    This paper presents a method for сomputed tomography imaging analysis for disease diagnosis, extending and fine-tuning a previously trained network to generate embedding vectors. A KNeighborsClassifier trained on produced embeddings achieved an accuracy of 0.987.
  • Ескіз
    ДокументВідкритий доступ
    Multi-class classification of pulmonary diseases using computer tomography images
    (КПІ ім. Ігоря Сікорського, 2023) Smilianets, F.; Finogenov, O.
    This paper examines approaches to classifying pulmonary diseases using neural networks. A modification of an existing neural network architecture for multi-class classification based on CT scans is proposed. The proposed architecture distinguishes between coronavirus pneumonia, non-hospital pneumonia, and healthy lungs. The training procedure of the proposed neural network, final parameters, and classification results are described. Conclusions are drawn regarding the potential applications of the proposed modification.