Przeglądaj {{ collection }} wg Autor "Szczepaniak, Piotr S."

Teraz wyświetlane 1 - 4 z 4

Case-Based Reasoning for Pattern Recognition using Granular Information Generated by Active Hypercontours
(Wydawnictwo Politechniki Łódzkiej, 2021) Szczepaniak, Piotr S.
This paper presents a novel extension of the case-based reasoning (CBR) technique. In the proposed method, a case is defined using the concept of multidimensional information granule created by the active hypercontour. The granularity of information is expected to be justifiable. The paper explains the main principles of the method and discusses its usefulness with reference to the pattern recognition problem.
Klasyczne i rozmyte bazy danych : modele, zapytania i podsumowania
(Akademicka Oficyna Wydawnicza EXIT, Warszawa, 2008) Zadrożny, Sławomir; Myszkorowski, Krzysztof; Szczepaniak, Piotr S.; Piecha, Jan
Rule Extraction from Active Contour Classifiers
(Wydawnictwo Politechniki Łódzkiej, 2018) Szczepaniak, Piotr S.; Pierścieniewski, Łukasz
In this paper, the idea of rule extraciton from active contour classifiers is presented. The concepts are new in relation to active contour approach. The problem is illustrated by examples having roots in technical diagnosis and in analysis of content of images.
Synergy of Convolutional Neural Networks and Geometric Active Contours
(Wydawnictwo Politechniki Łódzkiej, 2021) Tomczyk, Arkadiusz; Pankiv, Oleksandr; Szczepaniak, Piotr S.
Hybrid approach to machine learning techniques could potentially provide improvements in image segmentation results. In this paper, a model of cooperation of convolutional neural networks and geometric active contours is proposed and developed. The novelty of the approach lies in combining deep neural networks and active contour model in order to improve CNN output results. The method is examined on the image segmentation task and applied to the detection and extraction of nuclei of HL60 cell line. The model had been tested on both 2-D and 3-D images. Because of feature learning characteristics of convolutional neural networks, the proposed solution should perform well in multiple scenarios and can be considered generic.