CPClaudiu Ionut Popirlan

Claudiu Ionut Popirlan

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Claudiu Ionut Popirlan has a total of 33 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Comparative Deep Learning Analysis for Asphalt Damage Detection Using Real and Range Imaging

    by Andrei-Gabriel Nascu, Cristina Popîrlan, Irina-Valentina Tudor, Daniela Dănciulescu, Mihaela Colhon, Cristina-Mihaela Tudorache, Constantin-Cristian Dinu, Gabriel Stoian, Claudiu-Ionuţ Popîrlan

  • LungAtten-Net: a dual-stage framework integrating attention and automated slice selection for lung cancer subtype classification

    by Sandra Jervas, Chinnu Jacob, Imthias Ahamed T Parambath, Resmi R, Cristina Popirlan, Claudiu-Ionut Popirlan, D Jude Hemanth

    Lung cancer remains one of the leading causes of death worldwide due to late detection and delays in biopsies, which impact treatment decisions. Different subtypes require various therapies, showing the need for automated, non-invasive nodule classification. Despite improvements in deep learning, slice selection in volumetric computed tomography (CT) scans still depends on manual extraction, making the process slow and inconsistent. Furthermore, effectively capturing both fine details and overall context across different subtypes is still a challenge. To address these issues, this study introduces a dual-phase deep learning pipeline. The first stage uses an optimized InceptionV3 model designed for automated lung organ slice selection in CT scans, which takes advantage of its multiscale feature extraction. A comprehensive ablation study comparing full-slice usage, random slice selection, and the proposed automated slice selection is conducted. The second stage processes the chosen slices with our proposed LungAtten-Net model, a classification network based on VGG16 that includes attention-refinement modules. This approach helps the model concentrate on the most relevant regions, which enhances its diagnostic performance. Evaluation on two publicly available datasets shows the method’s effectiveness. Experimental results on these datasets back up the effectiveness of the approach. The optimized InceptionV3 model shows better performance in automated lung slice selection, ensuring precise extraction of relevant areas. Meanwhile, LungAtten-Net outperforms baseline models, achieving an accuracy of 98.99% and 99.68%, along with superior precision, recall, F 1 score and AUC score. Model stability is further validated using 5-fold cross-validation and external dataset evaluation. In addition to standard performance assessments, we incorporate explainable AI methods, including gradient-weighted class activation mapping and local interpretable model-agnostic explanation, to produce both visual explanations and feature-level insights of the model’s decision-making process. Our approach improves CT scan analysis and improves classification of lung cancer subtypes, allowing early detection and better treatment decisions.

  • Hybrid denoising framework with deep learning and optimization for mixed noise removal in remote sensing images

    by A Josephine Atchaya, Cristina Popirlan, Claudiu-Lonut Popirlan, J Anitha

    Remote sensing images (RSIs) are often degraded by mixed noise, particularly salt-and-pepper noise (SPN) and white Gaussian noise (WGN), which adversely affect visual quality and analytical reliability. Traditional denoising methods struggle to suppress impulsive artifacts in the presence of mixed noise, especially in RGB images with varying luminance levels. To address this challenge, a novel hybrid denoising framework is proposed, integrating spatial filtering with deep learning techniques. The method first employs Contrast Limited Adaptive Histogram Equalization (CLAHE) to enhances contrast in shadowed and overexposed regions, followed by a Modified Decision-Based Unsymmetric Trimmed Median Filter (MDBUTMF) to suppress SPN noise while preserving edge fidelity. Gaussian Curvature Filtering (GCF) then refines geometrically significant regions and a Denoising Convolutional Neural Network (DnCNN) subsequently recovers fine textures and semantic details lost during spatial filtering. Additionally, the Adaptive Mayfly Optimization Algorithm (AMOA) is incorporated to dynamically tunes the filter parameters based on input image luminance variations. Experiments conducted on the UCM, WHU-RS19 and Landsat datasets demonstrate the superior performance of the proposed hybrid approach, achieving a peak PSNR of 32.80 dB, an SSIM of 0.888 and a minimum GMSD of 0.0537 under complex mixed noise conditions. The proposed framework generalizes effectively across diverse image content and noise levels, making it suitable for a wide range of remote sensing applications.


  • Autoregressive Model for Energy Consumption Time Series

    by Daniel-Costin Ebâncă, Claudiu-Ionuţ Popîrlan, Irina-Valentina Tudor, Cristina Popîrlan


  • Investigating User Characteristics and Behaviors on Social Media as Prequel for Fake News Detection

    by Claudiu-Ionuț Popîrlan, Cristina-Mihaela Tudorache, Irina-Valentina Tudor, Constantin-Cristian Dinu, Mihaela Colhon, Gabriel Stoian, Daniela Dănciulescu Computer, Cristina Popîrlan

    The rapid proliferation of social media platforms has led to a remarkable increase of fake news dissemination, creating a significant challenge for the integrity of online information. To analyze this issue, this study aims to investigate the relationship between user characteristics and behaviors on social media platforms as a prequel for the development of effective fake news detection strategies. For that purpose, we examined open access datasets available for Romanian internet users and selected several internet usage activities which we analyzed in relation to key demographic characteristics that lead to establishing user profiles prone to manipulations. By understanding the underlying factors that contribute to the spread of fake news, we can design targeted interventions to mitigate its impact and promote a more informed and responsible online environment.


  • Open Access

    Predicting the unemployment rate and energy poverty levels in selected European Union countries using an ARIMA-ARNN model

    by Claudiu Ionut Popirlan, Irina-Valentina Tudor, Cristina Popirlan

    This article analyzes the correlation between energy poverty percentage and unemployment rate for four European countries, Bulgaria, Hungary, Romania and Slovakia, comparing the results with the European average. The time series extracted from the datasets were imported in a hybrid model, namely ARIMA-ARNN, generating predictions for the two variables in order to analyze their interconnectivity. The results obtained from the hybrid model suggest that unemployment rate and energy poverty percentage have comparable tendencies, being strongly correlated. The forecasts suggest that this correlation will be maintained in the future unless appropriate governmental policies are implemented in order to lower the impact of other aspects on energy poverty.


  • Open Access

    Influence of Certain Social Aspects on Most Likely Internet Usage Scenarios

    by Claudiu-Ionuţ Popîrlan, Irina-Valentina Tudor, Constantin-Cristian Dinu, Gabriel Stoian, Cristina Popîrlan, Daniela Dănciulescu

    This paper aims to analyze the influence that certain social factors (education and area of residence) have on the most likely scenarios people encounter in their online activities. Among the possible scenarios of using the internet, based on everyday individuals’ activities, we selected seeking information about health, goods, and services, taking online courses, internet banking, and participating in social networks. Using data acquired from international databases over the 2002–2020 period, we proposed five hypotheses and applied a multilinear regression model to the data collected for four European countries, namely, Bulgaria, Greece, Romania, and Slovenia. We have analyzed the degree of confirmation for all five hypotheses. The results provided a better understanding of the influence of the above-mentioned factors on the considered scenarios, allowing stakeholders to define and propose specific development policies.


  • A proposed multi-agent based platform for monitoring and control of Active Power Distribution Systems

    by Ionut-Adrian Cojoaca, Constantin Bulac, Claudiu-Ionut Popirlan

    Giving the current integration of large numbers of small distributed generation units connected to the medium and low voltage power distribution grid, may lead to multiple problems, that limits network’s capacity. To overcome those challenges, the concept of Smart Grid implements the required characteristics to ensure voltage stability, minimize power losses supply/demand balance, etcetera. Because of the time constraints, the decentralized control seems to be the solution, which can be achieved by the multi agent-system technology. To test the applications of multi-agent system in the monitoring and control of active power distribution systems, this paper proposes an intelligent platform architecture. The used technologies are presented, with the reasons behind the selections, resulting in an architecture composed of Matlab – MySQL – JADE. At the level of Matlab, the modeling and simulation modules are developed, under the form of graphical user interface. A study case is presented, who will refer to a 15 bus medium voltage power distribution grid, which is modelled with the graphical module, and analyzed with the real time simulation module.


  • Open Access

    Hybrid Model for Unemployment Impact on Social Life

    by Claudiu-Ionuţ Popîrlan, Irina-Valentina Tudor, Constantin-Cristian Dinu, Gabriel Stoian, Cristina Popîrlan, Daniela Dănciulescu

    In this paper, we want to examine how unemployment impacts social life, and, by using datasets from six European countries, we analyze the effect of unemployment on two of the main aspects of social life: social exclusion and life satisfaction. First, we predict unemployment rates using the Auto Regressive Integrated Moving Average (ARIMA) model and the results are further used in a linear regression model alongside social exclusion and life satisfaction data, thus obtaining the hybrid model. With the help of the point prediction method, we use the hybrid model to predict new values for the two aspects of social life for the upcoming three years and we analyze the results obtained in order to better understand their interconnection. The results suggest that unemployment has particularly adverse effects on the subjective perception of life satisfaction, furthermore increasing the social exclusion percentage.


  • Smart Storage and Grid Services Based on Removable Modular Batteries for EV

    by Leonardo-Geo Manescu, Denisa Rusinaru, Cosmin Buzatu, Lucian Mandache, Anca Bandoi, Catalina Sitnikov, Claudiu Popirlan, Carmen Maria Negrila, Claudiu George Bocean

    The wider use of renewable sources and of electric vehicles has to be supported by smart storage. Becoming modular and removable, battery charging and use differ temporally and spatially. This increases both the EV and the grid flexibility, makes it easier to purchase EVs, avoid useless reinforcement of the grid and postpone investment until recurring congestions occur. The EV owners can rent charged batteries for on board or domestic use. Aggregators can deploy removable batteries or include them in virtual power plants, virtual power lines, utility scale batteries in order to provide smart grid services, capacity firming of variable renewable sources. Better usability, care and recycling are achieved when battery management is done by professionals. All these concepts are first explained and then illustrated by numerical simulation.