This research presents a novel multi-exposure fusion model termed ColorMEF.In contrast to conventional approaches that rely primarily on image luminance data, ColorMEF integrates chromatic color information to augment the image fusion process.The incorporation of chromatic information enables ColorMEF to outperform existing models, as substantiated by evaluation metrics such as SSIM, DISTS, and VSI.Furthermore, the model is systematically trained using an end-to-end framework that optimizes the MEF-SSIM function based on the input data.ColorMEF achieves state-of-the-art performance on 4 of 5 conventional full-reference metrics: SSIM 0.9181 (+0.0092 versus the best prior), GMSD 0.0645 (-0.0043, 6.3%), DISTS 0.9266 (+0.0144), and VSI 0.9857 (+0.0025); on VIF ColorMEF is slightly lower (0.5224 vs 0.5394 for IFCNN).On the paper-introduced FSMEF-SSIM metric, ColorMEF ranks second (0.9435 vs 0.9465 for MEF-Net).These results indicate that coupling the chroma and luma during fusion improves structural fidelity and perceived quality across diverse scenes.

This study addresses a statistics-based testing procedure aimed at optimizing quality assurance for custom-made and made-to-measure (CM/M2M) clothing patterns.The procedure focuses on identifying machine-detectable issues in patterns before they progress to the expensive and time-intensive manual tailoring phase, thereby streamlining the testing cycle and reducing both time and costs.By analysing patterns for individuals with similar body measurements and applying statistical methods, the study identified potential design errors in pattern pieces based on measurable properties such as perimeters, areas, and contour-defining lines.Advanced statistical techniques, including residual analysis, Cook's distance, and Mahalanobis distance, were employed to detect outliers and pinpoint potential construction errors.Further analysis of line properties using predictive models-such as linear regression, random forests, generalized additive models (GAM), and rpart decision trees-revealed that a high frequency of outliers often correlates with construction anomalies.This research demonstrates that predictive modelling and outlier detection are effective tools for identifying errors in CM/M2M pattern construction, contributing to improved garment accuracy and production efficiency.

This study presents an analysis of the effects of noise on the performance of Bayesian neural networks, comparing Monte Carlo Dropout (MC-D), Bayes by Backpropagation (BBB), and Variational Inference (VI) algorithms.The research aims to identify the strengths and weaknesses of each method in order to guide practitioners in selecting the most suitable method for various applications.The performance of the methods are evaluated using uncertainty estimates, accuracy, and robustness.This study gives insights into which method works best in noisy data and how well they can estimate the noise in the data.The results show that Bayes backpropagation is the most robust method in noisy conditions, while Monte Carlo Dropout is the most accurate in noise-free conditions.Variational Inference is the most sensitive to noise, but can be used as an estimator of the noisiness of the data set

The hyperspectral unmixing method is an algorithm that extracts material (usually referred to as endmembers) data from hyperspectral data, along with their corresponding abundances.Due to the lower spatial resolution of hyperspectral sensor data compared to conventional cameras, each pixel is more likely to contain mixed information from multiple endmembers.In turn, mixed hyperspectral data is less valuable for use in research or predictive models.One of the problems in hyperspectral unmixing is the lack of openly available, field-collected datasets, particularly those from agricultural and other UAV-gathered sources.In turn, hyperspectral unmixing algorithms are rarely tested on this data type.This paper proposes a hyperspectral unmixing algorithm based on the U-Net network architecture to achieve more accurate unmixing results on existing and newly created hyperspectral unmixing datasets.The proposed model is fully unsupervised and is not limited by data shape and size.We also developed and shared a hyperspectral unmixing dataset derived from blueberry field data collected using a hyperspectral camera mounted on a Unmanned Aerial Vehicle (UAV).Compared to the state-of-the-art transformer-based unmixing model, our proposed algorithm achieved approximately 20% lower endmember RMSE and more than 50% lower reconstruction error values.

Global navigation satellite systems (GNSS) historically been known as one of the newest technologies since the 1970s.GNSS originally developed for military purposes in the USA (GPS -Global Position System).There are several satellite systems in the world.Satellites, International Research Base Stations, regional/national and local base stations form a permanent geodetic frame.Research on the size and shape of the Earth-planet, climate, sea, urban planning.In geodesy, a network of global positioning base stations makes it possible to assess the movements of continents, land plates at international level.GNSS is an important technology in navigation, logistics, economics, land surveying and other "geo" sectors.GNSS equipment/receivers and their manufacturers are applying new designs and electronics.Initially, GNSS instruments were used with single frequency signal reception, later expanding the number of GNSS signal channels to two frequencies.Such technological improvements nowadays improve the certainty, reliability and accuracythe overall quality of GNSS measurements.The GNSS base station enables the surveyors to determine coordinates with an accuracy of two centimeters in real time (RTK) and with an accuracy of five millimeters using the accumulated post-processing data.Various types of factors hamper GNSS measurements.The GNSS signal (radio wave) travels in airspace, in urban environments and is a physical parameter.Any obstacletree, building walls, and atmospheric effectmakes GNSS measurements less accurate.The GNSS signal must be strong and free from attenuation and suppression effects.This study develops GNSS models that show the comparison, certainty and reliability of GNSS measurements using different types of GNSS techniques.Evaluation of Latvian Global Positioning Reference Station Network -LatPos system measurements against a corresponding RTK solution method using Latvian Global Positioning Network geodetic point (G2 class).

The paper provides an overview of the digital humanities landscape in Lithuania, highlighting key initiatives, strategies, and resources that contribute to the field's development.It begins by defining the broad and evolving scope of digital humanities, examining its interdisciplinary nature and the challenges in establishing clear boundaries of the field.Admitting a major role of digital humanities in preserving cultural and national heritage, the paper overviews key initiatives in Lithuania, prominent heritage digitization platforms, along with significant language resources.The paper also presents several digital humanities' use cases, including offensive language identification and digital archives of Lithuanian literature.The conclusion reflects on future trends, emphasizing the increasing role of artificial intelligence and the continued digitization of cultural heritage in advancing digital humanities in Lithuania.