Anil Carie

Dr. Anil Carie


Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
30 March 2023 in Electronics
Verified
Journal Article
Open access
PDF file
BKBudati Anil Kumar
Budati Anil Kumar
SVSridhar Reddy Vulapula
Sridhar Reddy Vulapula
SSSyed Bilal Hussain Shah
Syed Bilal Hussain Shah
AAAnas Al-Tirawi
Anas Al-Tirawi
ACAnil Carie
Anil Carie
The hybrid cloud is a secure alternative for enterprises to exploit the benefits of cloud computing to overcome the privacy and security concerns of data in IoT networks. However, in hybrid cloud IoT, sensitive items such as keys in the private cloud can become compromised due to internal attacks. Once these keys are compromised, the encrypted data in the public cloud are no longer secure. This work proposes a secure multilevel privacy-protection scheme based on Generative Adversarial Networks (GAN) for hybrid cloud IoT. The scheme secures sensitive information in the private cloud against internal compromises. GAN is used to generate a mask with the input of sensory data-transformation values and a trapdoor key. GAN’s effectiveness is thoroughly assessed using Peak Signal-to-Noise Ratio (PSNR), computation time, retrieval time, and storage overhead frameworks. The obtained results reveal that the security scheme being proposed is found to require a negligible storage overhead and a 4% overhead for upload/retrieval compared to the existing works.

28 May 2026 in Frontiers in Blockchain
Verified
Journal Article
Open access
PDF file
IOIsaac Osei
Isaac Osei
ACAnil Carie
Anil Carie
SASatish Anamalamudi
Satish Anamalamudi
DBDennis Opoku Boadu
Dennis Opoku Boadu
Purpose: This study investigates the drivers of cryptocurrency adoption in Ghana, an emerging economy where adoption patterns remain underexplored. It seeks to explain both the linear relationships and complex configurational pathways that shape user behaviour in digital financial systems. Design/methodology/approach: A dual-method approach was employed, combining Partial Least Squares Structural Equation Modelling (PLS-SEM) and fuzzy-set Qualitative Comparative Analysis (fsQCA). PLS-SEM was used to examine net effects among constructs, while fsQCA identified multiple causal configurations leading to adoption. Data were collected from 652 cryptocurrency users in Ghana through a structured online survey. Findings: The PLS-SEM results indicate that facilitating conditions and intention to use are the most direct predictors of actual usage. However, several relationships exhibit counterintuitive effects, including negative influences of financial literacy and perceived usefulness on attitude, and a negative direct effect of behavioural intention on usage. The fsQCA results complement these findings by revealing multiple alternative pathways to adoption, including configurations in which users adopt despite low intention, driven by enabling conditions or perceived opportunity. These results highlight the non-linear, asymmetric, and context-dependent nature of cryptocurrency adoption. Originality: This study contributes to the technology adoption literature by integrating Technology Acceptance Model and UTAUT constructs within a combined PLS-SEM and fsQCA framework. It extends existing models by demonstrating that adoption behaviour in high-risk financial contexts cannot be fully explained through linear assumptions alone, thereby introducing a configurational perspective to cryptocurrency adoption research in emerging economies. Practical implications: The findings suggest that policymakers and fintech developers should prioritise infrastructural support, user education, and context-sensitive strategies. Enhancing facilitating conditions and financial literacy, while reducing barriers to use, may be more effective than focusing solely on attitudinal change.

25 February 2026 in Frontiers in Human Dynamics
Verified
Journal Article
Open access
PDF file
IOIsaac Osei
Isaac Osei
ACAnil Carie
Anil Carie
Human performance, psychological well-being, and behavioral adaptation in isolated, confined, and extreme (ICE) environments—such as intensive care units, remote healthcare settings, disaster zones, and space-analog missions—are increasingly mediated through digital and interactive technologies. Understanding how users perceive, experience, and sustain engagement with such systems requires methodological approaches capable of modeling complex psychosocial constructs under constrained conditions. This paper positions Partial Least Squares Structural Equation Modelling (PLS-SEM) as a robust and practical analytical framework for Human–Computer Interaction (HCI) research in ICE contexts. We review the methodological limitations of traditional statistical approaches and demonstrate how PLS-SEM enables the simultaneous modeling of latent experiential, cognitive, and behavioral variables central to technology use in extreme environments. Using an illustrative case involving a mobile health (mHealth) system designed for remote and resource-constrained settings, we examine how user experience, user satisfaction, perceived usefulness, perceived ease of use, and system quality influence behavioral intention. The findings indicate that experiential and affective factors—particularly user experience and satisfaction—play a dominant role in shaping continued usage intentions, while functional attributes exert their influence primarily through indirect pathways. By reframing PLS-SEM as an enabling methodology for ICE research, this paper contributes a theoretically grounded and practically applicable roadmap for researchers investigating human–technology interaction in extreme environments.

20 March 2026 in Frontiers in Human Dynamics
Verified
Journal Article
Open access
PDF file
IOIsaac Osei
Isaac Osei
ACAnil Carie
Anil Carie
LKLakshmi Prasanna Kanithi
Lakshmi Prasanna Kanithi
DBDennis Opoku Boadu
Dennis Opoku Boadu
Human activity is increasingly extending into environments marked by isolation, confinement, and extreme conditions, including long-duration space missions, polar research stations, intensive care units, and other high-risk settings. In these contexts, individuals must sustain performance and wellbeing under persistent cognitive, emotional, and social strain. Artificial intelligence (AI) systems are now deeply embedded in such environments, supporting decision making, monitoring, training, and, in some cases, psychological wellbeing. Yet research on AI in extreme settings has largely emphasized technical performance and automation, with comparatively limited attention to the lived experience of sustained Human–AI Interaction (HAI). This perspective paper argues that isolated, confined, or extreme (ICE) environments represent a uniquely revealing context for examining HAI. The psychological pressures characteristic of ICE settings—such as prolonged isolation, cognitive fatigue, stress, and high consequences of error—fundamentally shape how humans perceive, trust, and rely on AI systems. Drawing on interdisciplinary literature from human factors, psychology, and AI research, the paper conceptualizes ICE environments as a stress test for HAI, where issues of trust calibration, autonomy, transparency, and social attribution are amplified. Rather than treating AI solely as a decision aid, this perspective highlights how AI systems in ICE contexts may function as cognitive partners, social surrogates, or perceived teammates. The paper concludes by outlining key implications for the design, evaluation, and governance of AI systems intended for extreme environments, emphasizing the need for interaction-centered approaches that prioritize human experience alongside technical performance.

1 January 2026 in Intelligent Automation and Soft Computing
Verified
Journal Article
Open access
PDF file
CBChanchal Biswas
Chanchal Biswas
NMNaveenkumar Mahamkali
Naveenkumar Mahamkali
SASatish Anamalamudi
Satish Anamalamudi
BMBhaskar Marapelli
Bhaskar Marapelli
ACAnil Carie
Anil Carie
Effective feature extraction is problematic due to subtle differences in skin texture, color, and shape. Moreover, imbalanced datasets are common in medical image analysis, which complicates classification by biasing models toward dominant classes, causing overfitting. The proposed approach is a hybrid one comprising transfer learning and self-attention. We have employed ResNet-50 pre-trained model as a feature extractor for dermatoscopic images. Max Pooling and Global Average Pooling were used to focus on the important patterns and reduce irrelevant background information. He normal kernel and L2-regularization initialization are applied to emphasize salient patterns and suppress irrelevant background information. The module of self-attention comes after the max-pooling and before the global average-pooling. This method allows for efficient modeling of lesion relationships across the entire globe in a dual pooling strategy. We have used particle swarm optimization (PSO) to tune the hyperparameters. Optimization methods, such as early stopping, ReduceLROnPlateau, dropout, and 10-fold cross-validation, help to guarantee strong convergence and thus avoid overfitting. In addition, we have used data augmentation and transfer learning to overcome the imbalance in the datasets. This will help to generalize across different lesion types. The proposed model achieves an accuracy of 97.77%, precision of 98.00%, recall of 97.17%, and F1-score of 97.45% on the HAM10000 dataset. To assess the robustness and generalization ability of the proposed solution, we compared using the ISIC-2019, ISIC-2019-2020, and ISIC datasets, which include both binary and multi-class classification. The proposed solution has also achieved better performance compared to several recent research works conducted to solve the challenges regarding skin cancer feature extraction and hence improve cancer detection accuracy. We have demonstrated that the proposed framework effectively focuses on clinically relevant lesion regions using Grad-CAM as an explainable AI (XAI) technique.

1 January 2026 in Intelligent Automation and Soft Computing
Verified
Journal Article
Open access
PDF file
ACAnil Carie
Anil Carie
SMSanjeev Kumar Makala
Sanjeev Kumar Makala
SASatish Anamalamudi
Satish Anamalamudi
STShaik Teena
Shaik Teena
ADAwadhesh Dixit
Awadhesh Dixit
BMBhaskar Marapelli
Bhaskar Marapelli
Connected Vehicle-to-Everything (C-V2X) networks rely on Basic Safety Messages (BSMs) for cooperative awareness, but authenticated vehicles can still transmit falsified kinematic data—an insider attack that cryptographic authentication cannot prevent. Existing misbehavior detection systems (MDSs) achieve high detection rates on static benchmarks yet provide no formal guarantee that honest behaviour is a rational vehicle’s dominant strategy. We present a four-layer hierarchical trust architecture for 5G New Radio (NR) C-V2X that integrates edge AI detection, cloud large language model (LLM)-driven weight calibration, game-theoretic conviction, and ledger-anchored payoff tracking. A Nash equilibrium gate, calibrated with a wide margin above the empirically observed honest-vehicle suspicion ceiling, achieves zero observed false positives across 1584 honest phase-3 instances (one-sided 95% Clopper–Pearson upper bound: 0.19% per instance), without requiring any assumption about the honest-score distribution’s shape. A clawback mechanism makes the expected payoff of attacking strictly negative for all tested temptation levels. Evaluation on NS3 with 3GPP Rel-18 channel models across nine attack types—including rational adversaries with heterogeneous temptation Ti∼U(0.5,2.5)—yields a detection rate (DR) of 86.3% with false-positive rate (FPR) of 0% observed across 1584 honest phase-3 instances. This zero-FPR result holds for the evaluated single-UAV configuration at N≤50 vehicles per zone; at N=100, mean FPR rises to 4.9% (range 0%–13.5% across seeds) due to coverage-boundary effects, and multi-UAV sectorisation to extend this range remains unvalidated futurework. All 95 rational attackers (N=30, seeds 1, 2, 3, 5, 6) chose cooperation, producing a honesty premium of Π=+39.2 Trust Credits (positive in every seed, range +16.4 to +55.4 TC). A five-arm ablation isolates the Nash gate’s and clawback’s independent contributions to this outcome. Investigating cloud LLM calibration’s contribution to detection rate, we identified and corrected a learning-rate implementation fault that had suppressed nearly all of the LLM’s recommended adjustments throughout the original evaluation; under the corrected configuration, calibration produces a statistically significant detection-rate improvement, confirmed with a live cloud-unreachable control (+5.3 percentage points, p=0.034), reported alongside the original finding for transparency. To our knowledge, this is the first V2X misbehavior detection system to provide formal incentive compatibility under heterogeneous temptation for economically-rational attackers under the specified payoff model; this claim does not extend to attackers pursuing non-economic objectives, collusion, Sybil behaviour, or calibration-pipeline poisoning, which remain open problems.