Contrast Features Research Articles

Can Multimodal Large Language Models Enhance Performance Benefits Among Higher Education Students? An Investigation Based on the Task–Technology Fit Theory and the Artificial Intelligence Device Use Acceptance Model

The current study highlights the potential of multimodal large language models (MLLMs) to transform higher education by identifying key factors influencing their acceptance and effectiveness. Aligning technology features with educational needs can enhance student engagement and learning outcomes. The study examined the role of MLLMs in enhancing performance benefits among higher education students, using the task–technology fit (T-TF) theory and the artificial intelligence device use acceptance (AIDUA) model. A structured questionnaire was used to assess the perceptions of 550 Saudi university students from various academic disciplines. The data were analyzed via structural equation modeling (SEM) using SmartPLS 3.0. The findings revealed that social influence negatively affected effort expectancy regarding MLLMs and that hedonic motivation was also negatively related to effort expectancy. The findings revealed that social influence and hedonic motivation negatively affected effort expectancy for MLLMs. Effort expectancy was also negatively associated with T-TF in the learning context. In contrast, task and technology characteristics significantly influenced T-TF, which positively impacted both performance benefits and the willingness to accept the use of MLLMs. A strong relationship was found between adoption willingness and improved performance benefits. The findings empower educators to strategically enhance MLLMs adoption strategically, driving transformative learning outcomes.

Contrastive feature hierarchies and Germanic phonology

Abstract I discuss an analysis of changes in the Scandinavian runic alphabet, or futhark, by Jørgen Rischel (1966). Rischel’s article accounts for some puzzling changes in the futhark by employing contrastive feature hierarchies represented as branching trees. Feature hierarchies can be traced back to the work of Roman Jakobson and his colleagues. They enjoyed a brief period of prominence in the 1950s and 1960s, but then disappeared from mainstream phonological theory. However, they were employed in a number of interesting studies of Germanic and other languages whose insights we can still profit from today. The goal of this paper is to bring attention to this largely forgotten approach to phonological analysis, and to spell out the principles that underlie it.

St-RegSeg: an unsupervised registration-based framework for multimodal magnetic resonance imaging stroke lesion segmentation.

Stroke is one of the leading causes of disability and death worldwide. Ischemic stroke accounts for 75-90% of all stroke incidents. Assessing the size and location of the stroke lesion is crucial for treatment decisions, especially those related to urgent vascular reconstruction surgery. Magnetic resonance imaging (MRI) offers excellent soft tissue contrast and multimodal imaging characteristics, which can reflect changes in the physiological functions of brain soft tissues in patients with ischemic stroke. However, using deep learning (DL) techniques for MRI segmentation of stroke lesions still faces many challenges. On the one hand, single-modal segmentation models cannot effectively integrate multimodal information; on the other hand, there is a semantic content drift between multimodal stroke MRI images, leading to lower accuracy in subsequent multimodal image segmentation. To address these issues, we aimed to propose the stroke unsupervised registration and segmentation (St-RegSeg) framework. The St-RegSeg framework integrates an unsupervised registration model, ConvNXMorph, and a segmentation model, nnUNet-v2, enabling both registration and segmentation of multimodal MRI images. The St-RegSeg framework was evaluated on the ISLES'22 dataset from three centers. The St-RegSeg framework demonstrated significant improvements in performance metrics and computational efficiency. Compared to advanced normalization tools (ANTs) [symmetric normalization (SyN)] + nnUNet-v2, the St-RegSeg framework improved the Dice similarity coefficient (DSC) by 25.31% in the registration phase, reduced mean squared error (MSE) by 17.36%, increased normalized cross-correlation (NCC) by 16.06%, and enhanced mutual information (MI) by 17.09%. Additionally, in the segmentation phase, it increased the DSC by 0.84%, and the overall inference speed was increased by 40.91 times. Compared to the suboptimal TransMorph + nnUNet-v2, the St-RegSeg framework improved the DSC by 3.68% in the registration phase, reduced MSE by 8.91%, increased NCC by 8.49%, enhanced MI by 6.18%, and in the segmentation phase, it raised the DSC by 0.5%, with the overall inference speed increased by 2.13 times. The St-RegSeg framework provides a highly effective solution for the registration and segmentation of multimodal MRI images in ischemic stroke cases. Its performance metrics and computational efficiency significantly outperform existing methods, making it a promising tool for clinical applications. The code is open-sourced and available at: https://github.com/Cooper-Gu/St-RegSeg.

Streaming disasters on TikTok: Examining social mediated crisis communication, public engagement, and emotional responses during the 2023 Maui wildfire

Short video-sharing platforms, such as TikTok, play a significant role in allowing the public to experience natural disasters vicariously, share information, and coordinate peer-led disaster relief efforts. With emerging platforms like TikTok providing experientially immersive content, the role of digital storytelling in stimulating public engagement and emotions remains underexplored. Drawing from narrative persuasion theory and visual storytelling literature, the current study proposes an integrated framework to examine the storyteller, visual frame, and digital presentation characteristics of TikTok videos in predicting social-mediated public engagement and emotional reactions surrounding a recent natural disaster, the 2023 Maui Wildfire. Using visual content analysis and computational methods, the study analyzed 526 TikTok videos and 59,950 associated comments. Results showed that audiovisual vividness consistently predicted all three types of disaster engagement. In contrast, storyteller type, storytelling frame, and other presentation characteristics inconsistently predicted information consumption intention, information sharing intention, and emotional reactions on TikTok.

Class Hierarchy Aware Contrastive Feature Learning for Multigranularity SAR Target Recognition

Class Hierarchy Aware Contrastive Feature Learning for Multigranularity SAR Target Recognition

The Determinants of Earnings Management in Manufacturing Companies Listed on the Indonesia Stock Exchange

Profit control is an important issue in corporate governance arising from the conflict of interest between owners and managers. Weak supervision can facilitate earnings management practices by managers. This study aims to empirically examine the effect of firm size, institutional ownership, independent commissioners, the board of directors, and the audit committee on earnings management. Using the purposive sampling method, 42 manufacturing companies listed on the Indonesia Stock Exchange for the period 2021-2023 were selected as samples. Data analysis includes classical assumption test, model goodness test, and multiple linear regression analysis. The results showed that independent commissioners, board of directors, and institutional ownership had no significant effect on earnings management. In contrast, firm size and audit committee characteristics have a negative effect on earnings management. The findings conclude that corporate governance mechanisms have diverse roles in controlling earnings management practices. This study suggests strengthening the role of the audit committee and increasing transparency in large companies to mitigate earnings management. The implications of this study emphasize the importance of evaluating and improving corporate governance mechanisms to improve the quality of financial reporting

MVE-Net: A label-free microscopic image visual enhancement network via mRetinex and nonreference loss guidance

Label-free microscopic cell image analysis (segmentation, detection, counting, e.g.) is elementary for unravelling the biological functions of cells and their organelles. However, low contrast, darker brightness, background inhomogeneous, and weak edges of cells cause challenges in subsequent cell image analysis processes. To address these challenges, a Microscopic Visual Enhancement Network (MVE-Net) is proposed to improve microscopic visual effects through pre-enhancement and enhancement processes. In the pre-enhancement stage, to overcome the difficulty of acquiring paired or unpaired images for training, the mRetinex block is proposed to guide the pre-enhancement network to image contrast, cell details, and structural features. Furthermore, a multi-scale extraction module is employed to extract and fuse cell texture and structural features from the pre-enhanced images at various scales, guiding the generator training. In the enhancement stage, a nonreference loss block is designed, incorporating spatial consistency, uneven illumination smoothness, and exposure adjustment loss terms, to further enhance the contrast between cells and the background, smooth the inhomogeneous background, and adjust overall image brightness, thereby guiding the generator's enhancement process and improving the visual effect of microscopic images. Experiments on the LIVECell and PNT1A datasets demonstrate that MVE-Net outperforms state-of-the-art image enhancement methods, significantly improving image contrast, brightness, cell detail, and structural features without the need for paired or unpaired reference standard images for training.

Titanite as an indicator of granite fertility and gold mineralization in the Xiaoqinling gold province, North China Craton

The Xiaoqinling gold province, located in the southern margin of the North China Craton (NCC), is the second largest gold-enriched region in China. In this region, the Mesozoic Huashan (HS) and Wenyu (WY) plutons are the major magmatic intrusions coeval with gold mineralization, although they show contrasting characteristics in the distribution of gold. In this study, we use geochemical features of titanite determined by LA-ICP-MS and EPMA analyses and elemental mapping to decipher the mechanisms that led to the difference in gold enrichment related to the two plutons. Titanite from the Wenyu granitic pluton exhibits significantly higher (La/Sm)N, (La/Yb)N, ΣLREE/ΣHREE ratios, and ΣREE concentration and slightly higher (Gd/Yb)N values than those of the Huashan Pluton, suggesting that the Wenyu pluton might have experienced more complex magmatic evolution, widespread hydrothermal alteration, and higher silica activity in the melt than the Huashan pluton. The titanite grains from the Huashan pluton show higher (Nb/Ta)N and (Lu/Hf)N values and significantly lower Zr concentration than those of the Wenyu pluton. The titanite grains from the Wenyu pluton show higher vanadium and gallium concentrations and Fe/Al ratio than those of the Huashan pluton, indicating comparatively higher fo2. Furthermore, the titanite grains from Wenyu pluton indicate higher water content in the magma. In addition, magma mingling and magmatic hydrothermal fluids derived from the crust/mantle are critical sources for ore-forming materials. These results suggest that the Wenyu pluton is more conducive to gold migration and enrichment than the Huashan pluton.

The Importance of Pocket Parks on Air Quality: A Comparative Approach

Population growth, increasing construction, and impervious surfaces are among the factors leading to the deterioration of the natural balance. Increasing population density has led to a gradual decrease in green areas for living spaces, transport routes and car parks. In addition, the rate of urbanisation and related environmental problems are one of the biggest threats to the sustainability of open green spaces. Accurate identification of problems, such as air, water, and soil pollution is critical to addressing current environmental problems and preventing future problems. Air pollution poses a significant threat to the global community and has sequential impacts on health systems, ecosystems and economies in both developed and developing countries. Cities are home to 50 per cent of the world's population and are expected to reach 70 per cent by 2050. The rapid urbanisation process leads to the transformation of natural and semi-natural landscapes into impervious surfaces and increased heat absorption rates. This study focuses on Jumeirah Island in Dubai. This artificial island is home to many businesses that support social life. Each café in these areas is located with public green areas in front of it and its own open spaces. In the study, 40 measurement points were established in two contrasting environmental conditions and air quality measurements were made manually. The stations are divided into blue and red stations and have contrasting characteristics. While the red measurement points have more green areas, water elements, qualified landscape texture, the blue measurement points have these characters to a lesser extent. Thanks to this comparison, the effect of vegetation texture and pocket parks on air quality in the study area was analysed. As a result of the analyses, the air quality in both areas was found to be good and acceptable, while the results at the red measurement points were found to be of better quality than the blue measurement stations. As a result of the study, the effect of water elements, qualified landscape design and vegetative texture on air quality has been proved with numerical data.

MORPHOLOGICAL AND ANATOMICAL RESEARCH OF COMMON MYRTLE LEAVES (MYRTUS COMMUNIS L.)

This paper presents the results of morphological-anatomical study of common myrtle leaves diagnostically significant features (Myrtus communis L.) harvested in the Nikitsky Botanical Garden (Republic of Crimea, Russia). Microscopic analysis of common myrtle leaf petioles was carried out and diagnostic features of dominant biologically active сompounds and tissues luminescence of this plant species leaf petioles were revealed. Morphology and anatomy of common myrtle leaves petioles, luminescence of their tissues were analyzed and summarized and additional diagnostic features of this plant object were revealed. Morphological and anatomical study of common myrtle leaves (Myrtus communis L.) was carried out in accordance with the OFS "Leaves" and "Technique of microscopic and microchemical study of medicinal plant raw materials and medicinal plant preparations" GF RF XIV edition using light microscopy in transmitted and polarized light, as well as luminescence microscopy, which has not been previously studied for this plant. The results of the study concluded that when comparing the luminescence of whole essential oils of myrtle, impure eucalyptus and their components provides a luminescence diagnosis due to the difference in luminescence color; 1,8-cineole is not a diagnostic element in morphological and anatomical luminescence for common myrtle due to the lack of obvious luminescence; euvimal-1 can be used as a diagnostic and luminescence as it has specific luminescence in the form of pink color; the peculiarity of pigmentation of the epithelium of the schizogenic receptacle of the leaf of common myrtle was revealed and the characteristic of its luminescence is species specific for myrtle and can be used as a diagnostic feature in contrast to the impurity raw material of eucalyptus. The results obtained during the study can be further used in the development of the section of the draft pharmacopoeial article “Microscopic features” and other regulatory documentation for a new type of medicinal plant raw material - “Myrtle leaves”, as well as in fundamental botany in the study of species of the family Myrtaceae.

Swill and Pig Manure Substrates Differentially Affected Transcriptome and Metabolome of the Black Soldier Fly Larvae.

Black soldier fly larvae (Hermetia illucens) (BSFL) are insect larvae with significant ecological and economic value. This study aims to investigate whether swill and manure had any effects on the transcriptome and metabolome of BSFL. Through high-throughput transcriptome sequencing, we found that larvae fed with swill exhibited higher levels of gene expression, especially with the upregulation of genes related to energy metabolism, amino acid metabolism, and redox reactions. Metabolomics analysis showed a significant increase in energy metabolism-related metabolites, such as organic acids and amino acids, in the swill-fed larvae. In contrast, gene expression and metabolic characteristics in the pig manure-fed group indicated a higher stress response, with relevant genes and metabolites (such as short-chain fatty acids and antioxidants) showing significant upregulation. GO and KEGG enrichment analyses further supported these results, suggesting that swill promotes larval growth and metabolism, whereas pig manure induces the activation of stress response mechanisms. These findings offer clear molecular and physiological insights into the optimization of substrate selection for enhancing the performance of BSFL in waste management.

Neuroendocrine tumors of the stomach and pancreas. Possibilities of application of radiomics. Problems and solutions.

Background. Radiomics is promising tool contributing to the diagnosis and treatment of neuroendocrine neoplasms of various localizations. This method is often used for differential diagnosis of gastrointestinal neoplasms from other neoplasms of this localization.Aim. To demonstrate difficulties in imaging of gastrointestinal neuroendocrine tumors, as well as possibilities of radiomics in differential diagnosis.Materials and methods. Data of 12 patients with morphologically confirmed gastric neoplasms (6 - with neuroendocrine tumor, 6 - with adenocarcinoma) and data of 22 patients with morphologically confirmed pancreatic neoplasms (11 - with neuroendocrine tumor, 11 - with adenocarcinoma) who underwent CT examination of abdominal cavity organs with intravenous contrast before treatment at the Russian Scientific Center of Roentgenoradiology were used in this study. Radiomic features were extracted from the area of manually segmented gastric and pancreatic tumors in the native phase of the CT scans. Processing of the features and statistical analysis were performed using Microsoft Office Excel and free open-source software development environment for the R programming language - R-Studio.Results. Typical and atypical imaging features of gastric and pancreatic neuroendocrine tumors, peculiarities of contrast, localization and structure of neoplasms were demonstrated on the examples of CT scans. There were 15 radiomic features that were statistically significantly different in patients with neuroendocrine tumors and gastric adenocarcinomas. In the group of patients with neuroendocrine tumors and pancreatic adenocarcinomas, statistically significant differences were observed in 14 radiomic features.Conclusion. Neuroendocrine tumors of the stomach and pancreas are rare neoplasms that are not only clinically silent in most cases, but also difficult to visualize due to their small size and poor contrast characteristics. Texture analysis is a promising tool for the differential diagnosis of gastrointestinal neoplasms from other neoplasms of this localization, especially considering the difficulties of biopsy.

Dynamics of bubble rising in extremely high and low viscous fluids

Bubble dynamics is a captivating field of study due to its complexity and wide-ranging applications. The present research focuses on bubble behavior in fluids with very high and low viscosities. Our study also focuses on understanding the effect of gas flow rate and needle diameter on bubble size. This research is essential because of scientific disagreement in this area, and our revisit adds new findings to the problem. We perform experiments in liquids with extreme viscosity differences using needles ranging from 0.84 mm to 1.54 mm and gas flow rates from 5 ml/min to 80 ml/min. In low-viscosity fluid, bubble necking time is constant with the flow rate, and secondary necking leads to daughter bubble formation. In high-viscosity fluid, bubble formation and rising differ, with smaller bubbles sliding over larger ones and merging symmetrically along their axis. The symmetry of contact decides sliding or coalescence. We also noticed that bubbles formed pairs (doublets) when the gas flow rate decreased with increasing needle size. The consecutive bubble diameter in both fluids is explored. The contrasting characteristics in low-viscosity environments (such as deionized water) and high-viscosity environments (like glycerol) highlight their respective monotonic and non-uniform flow behaviors.

High-resolution gene expression atlases of two contrasting major Greek olive (Olea europaea L.) tree cultivars for oil and table olive production.

Olive (Olea europea L.) is one of the most economically important tree crops worldwide, especially for the countries in the Mediterranean basin. Given the economic and nutritional importance of the crop for olive oil and drupe production, we generated transcriptional atlases for the Greek olive cultivars "Chondrolia Chalkidikis" and "Koroneiki" which have contrasting characteristics in terms of fruit development, oil production properties, and use. Our analysis involved 14 different organs, tissue types, and developmental stages, including young and mature leaves, young and mature shoots, open and closed flowers, young and mature fruits (epicarp plus mesocarp), young and mature endocarps, stalks, as well as roots. The developed gene expression atlases and the associated resources offer a comprehensive insight into comparative gene expression patterns across several organs and tissue types between significant olive tree cultivars. The comparative analyses presented in this work between the "Koroneiki" cultivar, which performs better in olive oil production, and the "Chondrolia Chalkidikis," which grows larger fruits, will be essential for understanding the molecular mechanisms underlying olive oil production and fruit shape and size development. The developed resource is also expected to support functional genomics and molecular breeding efforts to enhance crop quality and productivity in olive trees. The transcriptome data were generated using paired-end Illumina Next-Generation Sequencing technologies. The sequencing yielded approximately 13 million reads per sample for "Chondrolia Chalkidikis" and around 24 million reads per sample for "Koroneiki." The transcriptomes were comparatively analyzed to reveal tissue-specific and differentially expressed genes and co-expression gene modules within and between cultivars. The comparative analysis unveiled tissue-specific and differentially expressed genes within and between cultivars. Hierarchical gene clustering revealed intra- and inter-cultivar expression patterns, particularly for the endocarp and fruit tissues relevant to olive oil production and fruit development. Additionally, genes associated with oil production and fruit size/shape development, including those in fatty acid metabolism and developmental regulation, were identified. To facilitate accessibility, the GrOlivedb (www.GrOlivedb.com) database was developed, housing the comprehensive transcriptomic data for all of the analyzed organs and tissue types per cultivar. This resource will be a useful molecular tool for future breeding studies in olive oil production and fruit development and a valuable resource for crop improvement.

Emotion-Semantic-Aware Dual Contrastive Learning for Epistemic Emotion Identification of Learner-Generated Reviews in MOOCs.

Identifying the epistemic emotions of learner-generated reviews in massive open online courses (MOOCs) can help instructors provide adaptive guidance and interventions for learners. The epistemic emotion identification task is a fine-grained identification task that contains multiple categories of emotions arising during the learning process. Previous studies only consider emotional or semantic information within the review texts alone, which leads to insufficient feature representation. In addition, some categories of epistemic emotions are ambiguously distributed in feature space, making them hard to be distinguished. In this article, we present an emotion-semantic-aware dual contrastive learning (ES-DCL) approach to tackle these issues. In order to learn sufficient feature representation, implicit semantic features and human-interpretable emotional features are, respectively, extracted from two different views to form complementary emotional-semantic features. On this basis, by leveraging the experience of domain experts and the input emotional-semantic features, two types of contrastive losses (label contrastive loss and feature contrastive loss) are formulated. They are designed to train the discriminative distribution of emotional-semantic features in the sample space and to solve the anisotropy problem between different categories of epistemic emotions. The proposed ES-DCL is compared with 11 other baseline models on four different disciplinary MOOCs review datasets. Extensive experimental results show that our approach improves the performance of epistemic emotion identification, and significantly outperforms state-of-the-art deep learning-based methods in learning more discriminative sentence representations.

Denoising low-field MR images with a deep learning algorithm based on simulated data from easily accessible open-source software

In this study, we introduce a denoising method aimed at improving the contrast ratio in low-field MRI (LFMRI) using an advanced 3D deep convolutional residual network model. Our approach employs synthetic brain imaging datasets that closely mimic the contrast and noise characteristics of LFMRI scans, addressing the limitation of available in-vivo LFMRI datasets for training deep learning models. In the simulation data, the Relative Contrast Ratio (RCR) increased, and similar improvements were observed in the in-vivo data across different imaging conditions. Comparative evaluations demonstrate that our model performs better than the widely used non-deep learning method, BM4D, in enhancing RCR and maintaining high spatial frequency components in in-vivo data.

Assessment of interfering/synergistic effects in the adsorption between polar and non-polar VOCs on a commercial biomass-based microporous carbon

This research has been carried out to investigate unique relationships in adsorption behavior between polar and non-polar volatile organic compounds (VOCs: formaldehyde (FA) versus toluene) using commercial macadamia nutshell (MNS)-based microporous activated carbon (i.e., Procarb-900: namely, P900). The breakthrough (BT) volume, adsorption capacity, and partition coefficient of P900 are estimated for 100 ppm FA as a single component and as a binary phase with 100 ppm toluene. The contrasting features of adsorption (such as interfering/synergistic relationships) for VOC mixtures with different polarities are accounted for in terms of interaction between the key variables (e.g., pore size distribution, adsorbent particle size, surface element compositions, and sorbent bed mass). Accordingly, the powdered P900 (0.212–0.6 mm: 150 mg) exhibits an adsorption capacity of 5.7 mg g−1 and a partition coefficient of 0.19 mol kg−1 Pa−1 for single-phase FA at the 10% BT level. Interestingly, its FA adsorption performance is synergistically improved in the presence of toluene (e.g., > 150%) in the early stage of adsorption (e.g., 10% BT), although their competition reduced its performance at 99% BT. The apparent synergistic trend in the early BT stage may possibly reflect diffusion resistance of the adsorbent (e.g., small particle size and developed ultra-micropore structure) and natural attributes of FA (e.g., low affinity and smaller kinetic diameter). The overall results of this study are expected to offer a better understanding of the mechanisms underlying the interactions between the mixed VOC system and microporous adsorbents.

Electronic Structures of Late versus Early Transition Metal Imido Complexes from 15N‐NMR Signatures

AbstractImido ligand is a ubiquitous motif in organometallic chemistry, serving roles spanning from ancillary ligands to reactive sites. The nature of M=N bond is highly depended on the metal centres and their d‐electron configuration, with late transition metal (TM) imido complexes exhibiting contrasting features when compared to their early TM analogues. Envisioning to uncover general electronic descriptor for the nature of imido ligands, we computationally investigate the solid‐state 15N NMR signatures of late TM imido complexes with various central metals, geometries and d‐electron counts, and compare them against these of the corresponding early TM systems. The spectroscopic signatures are mostly driven by the presence of filled, π‐symmetry orbitals in late TM imido complexes, suggesting the development of high‐lying π(M=N) and low‐lying σ/σ*(M=N) orbitals. This contrasts with what is observed for the reported early TM systems, for which high‐lying σ‐type orbitals determine the NMR signature. Noteworthily, Ni‐ and Pd‐imido complexes with formal d10 configurations exhibit highly asymmetric nitrogen‐15 NMR signature with extremely deshielded principal components, because of the presence of filled, high‐lying antibonding π*(M=N) orbitals, consistent with their high reactivity. The sensitive response of 15N NMR signature to the nature of metal sites further highlights that chemical shift is a useful reactivity descriptor.

Multimodal biometric identification: leveraging convolutional neural network (CNN) architectures and fusion techniques with fingerprint and finger vein data.

Advancements in multimodal biometrics, which integrate multiple biometric traits, promise to enhance the accuracy and robustness of identification systems. This study focuses on improving multimodal biometric identification by using fingerprint and finger vein images as the primary traits. We utilized the "NUPT-FPV" dataset, which contains a substantial number of finger vein and fingerprint images, which significantly aided our research. Convolutional neural networks (CNNs), renowned for their efficacy in computer vision tasks, are used in our model to extract distinct discriminative features. Specifically, we incorporate three popular CNN architectures: ResNet, VGGNet, and DenseNet. We explore three fusion strategies used in security applications: early fusion, late fusion, and score-level fusion. Early fusion integrates raw images at the input layer of a single CNN, combining information at the initial stages. Late fusion, in contrast, merges features after individual learning from each CNN model. Score-level fusion employs weighted aggregation to combine scores from each modality, leveraging the complementary information they provide. We also use contrast limited adaptive histogram equalization (CLAHE) to enhance fingerprint contrast and vein pattern features, improving feature visibility and extraction. Our evaluation metrics include accuracy, equal error rate (EER), and ROC curves. The fusion of CNN architectures and enhancement methods shows promising performance in identifying multimodal biometrics, aiming to increase identification accuracy. The proposed model offers a reliable authentication system using multiple biometrics to verify identity.

Blind image quality assessment using Beltrami filter-based contrast features (BF-bCF) & LSTM network

ABSTRACT Advanced networks excel in Blind Image Quality Assessment (BIQA), however accurate estimation of quality scores remains challenging due to unrevealed features extricated under various distortions. This article presents a Beltrami filter-based Contrast feature and long short-term memory (BF-bCF & LSTM) framework for feature extraction, suitable for both authentic and synthetic image distortions. The framework comprises four modules: an edge preservation and enhancement module, a distortion-conscious module, a weight module, and a quality-predicting network. The edge enhancement module assists the distortion-conscious module by enhancing the overall quality of the input grayscale image. The distortion-conscious module uses a modified Difference of Gaussian (DoG) measure to mitigate the distortion and assign patch weights in the filtered image. The proposed BIQA framework assimilating LSTM effectively extricates significant components using principal component analysis (PCA) and predicts nearer scores. Experiments on synthetic and authentic datasets showed superior performance relating to SROCC and PLCC above 0.9.