Ajay Kumar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ajay Kumar has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Praneet Saurabh
Department of Computer Science and Engineering, Manipal University Jaipur, Jaipur, India
3 shared publications - Alpana Suman
Manipal University Jaipur, Jaipur, India
2 shared publications - Kedar Nath Singh
Department of Computer Science and Engineering, Lovely Professional University, Phagwara, India
2 shared publications - Vivek Bhardwaj
School of Computer Science and Engineering, Manipal University Jaipur, Jaipur, India
2 shared publications - Abhishek Singh Rathore
Department of Computer Science and Engineering, Shri Vaishnav Vidyapeeth Vishwavidyalaya, Indore, Madhya Pradesh, 453111, India
1 shared publication - Ajit Noonia
Department of Computer Science and Engineering, Manipal University Jaipur, Jaipur, India
1 shared publication
Latest Publications
- Open Access
- 14 November 2026
- Published by Springer Nature in Scientific Reports
Heuriskryptes framework for identifying covert channels in neural architectures
This paper introduces Heuriskryptes, a detection framework designed to identify AI-driven steganographic attacks within deep neural networks. The framework addresses the emerging threat of covert channels established within a model’s parameter pool. The evaluated attack vector, termed Stego-GCP, uses an orthogonal embedding to modify model weights with high stealth to leak data. To detect these channels, a specialized agent utilizes a Vision Transformer (ViT) architecture that treats the raw parameter space as an image, providing a self-attention-based mechanism that demonstrates consistent cross-architecture generalizability for efficient model auditing. The framework has been validated using a PooledCNN architecture on the Apple Disease and CIFAR-10 datasets. Results indicate that the carrier model retains a classification accuracy of 99.97%. Furthermore, the Heuriskryptes detector demonstrated high sensitivity, achieving 93.51% overall accuracy and maintaining a 90% detection rate specifically on the most challenging ambiguous, low-payload density scenarios ( $$\alpha < 0.002$$ ). This demonstrates highly competitive performance compared to evaluated frequency-domain anomaly detection baselines. Heuriskryptes serves as a reliable detection framework for verifying model integrity in collaborative and commercial AI environments. By focusing on the structural mathematical properties of the weight spectrum rather than activation-trigger patterns, this research provides a reliable defense against the unauthorized use of neural architectures as covert communication channels.- 14 November 2026
- Published by Springer Nature in Scientific Reports
Open AccessHeuriskryptes framework for identifying covert channels in neural architectures
This paper introduces Heuriskryptes, a detection framework designed to identify AI-driven steganographic attacks within deep neural networks. The framework addresses the emerging threat of covert channels established within a model’s parameter pool. The evaluated attack vector, termed Stego-GCP, uses an orthogonal embedding to modify model weights with high stealth to leak data. To detect these channels, a specialized agent utilizes a Vision Transformer (ViT) architecture that treats the raw parameter space as an image, providing a self-attention-based mechanism that demonstrates consistent cross-architecture generalizability for efficient model auditing. The framework has been validated using a PooledCNN architecture on the Apple Disease and CIFAR-10 datasets. Results indicate that the carrier model retains a classification accuracy of 99.97%. Furthermore, the Heuriskryptes detector demonstrated high sensitivity, achieving 93.51% overall accuracy and maintaining a 90% detection rate specifically on the most challenging ambiguous, low-payload density scenarios ( $$\alpha < 0.002$$ ). This demonstrates highly competitive performance compared to evaluated frequency-domain anomaly detection baselines. Heuriskryptes serves as a reliable detection framework for verifying model integrity in collaborative and commercial AI environments. By focusing on the structural mathematical properties of the weight spectrum rather than activation-trigger patterns, this research provides a reliable defense against the unauthorized use of neural architectures as covert communication channels. - Open Access
- 14 November 2026
- Published by Springer Nature in Discover Internet of Things
- Vol. 6 (1) , 272
A systematic review and conceptual hybrid framework for integrating blockchain and internet of medical things in secure healthcare systems
Internet of Medical Things (IoMT) deployments generate continuous streams of sensitive patient data, yet centralised storage remains exposed to single points of failure, unauthorised disclosure, and tampering. Blockchain offers an immutable, auditable substrate, but pairing it with resource-constrained devices raises concerns around throughput, latency, energy, and scalability. This paper presents a systematic literature review of peer-reviewed studies published between 2018 and 2025, selected under a PRISMA-aligned protocol across five databases, with all screening decisions double-rated (κ = 0.80). The review organises recurring technical gaps into a structured taxonomy, from which we derive a conceptual hybrid architecture combining an IoMT perception layer, an edge-gateway tier, a permissioned blockchain running Practical Byzantine Fault Tolerance (PBFT) consensus, and an Interplanetary File System (IPFS) off-chain store. The device–gateway key exchange is verified in ProVerif under a Dolev-Yao adversary, establishing session-key secrecy, injective agreement, and forward secrecy over unbounded sessions. Cryptographic costs measured on constrained hardware yield 35.7–141.4 ms commit latency and 707–2,798 transactions per second across a set of 4–31 validators, with an on-chain to off-chain storage ratio near 1:20. Cryptography accounts for only 2.8–10.7% of commit latency, confirming that network delay is the binding constraint. The contribution over prior surveys is a reproducible protocol, an explicit gap analysis, a unified notation and threat model, a machine-checked proof of authentication, and an analytical committee-sizing rule.
- 14 November 2026
- Published by Springer Nature in Discover Internet of Things
- Vol. 6 (1) , 272
Open AccessA systematic review and conceptual hybrid framework for integrating blockchain and internet of medical things in secure healthcare systems
Internet of Medical Things (IoMT) deployments generate continuous streams of sensitive patient data, yet centralised storage remains exposed to single points of failure, unauthorised disclosure, and tampering. Blockchain offers an immutable, auditable substrate, but pairing it with resource-constrained devices raises concerns around throughput, latency, energy, and scalability. This paper presents a systematic literature review of peer-reviewed studies published between 2018 and 2025, selected under a PRISMA-aligned protocol across five databases, with all screening decisions double-rated (κ = 0.80). The review organises recurring technical gaps into a structured taxonomy, from which we derive a conceptual hybrid architecture combining an IoMT perception layer, an edge-gateway tier, a permissioned blockchain running Practical Byzantine Fault Tolerance (PBFT) consensus, and an Interplanetary File System (IPFS) off-chain store. The device–gateway key exchange is verified in ProVerif under a Dolev-Yao adversary, establishing session-key secrecy, injective agreement, and forward secrecy over unbounded sessions. Cryptographic costs measured on constrained hardware yield 35.7–141.4 ms commit latency and 707–2,798 transactions per second across a set of 4–31 validators, with an on-chain to off-chain storage ratio near 1:20. Cryptography accounts for only 2.8–10.7% of commit latency, confirming that network delay is the binding constraint. The contribution over prior surveys is a reproducible protocol, an explicit gap analysis, a unified notation and threat model, a machine-checked proof of authentication, and an analytical committee-sizing rule.
- Open Access
- 25 September 2026
- Published by Springer Nature in International Journal of Computational Intelligence Systems
Reimagining Education: A Comparative Study of Artificial Intelligence and Large Language Model-based Learning Systems and Traditional Education
This study presents a comparative analysis of Artificial Intelligence (AI) and Large Language Model (LLM)-based learning systems versus traditional education methods, focusing on comprehension, retention, personalization, and critical thinking. With AI and LLM technologies transforming instructional design, it is essential to examine their pedagogical effectiveness and implications for higher education. A convergent mixed-methods approach was employed, combining quantitative and qualitative data collected from 350 students and faculty across different Indian academic institutions. A structured survey assessed perceptions of learning effectiveness, engagement, and assessment accuracy across both systems. Findings indicate that nearly half of the respondents viewed AI/LLM-based systems as more effective in enhancing comprehension and retention, while a majority recognized their strength in personalization and adaptive feedback. Traditional methods, however, remained valuable for fostering mentorship, ethical awareness, and collaborative learning. The study highlights the complementary strengths of AI-driven and traditional approaches, suggesting that hybrid learning models can maximize educational outcomes by integrating technological adaptability with human-centred pedagogy. The study contributes empirical evidence on the comparative effectiveness of AI/LLM-based and traditional learning approaches and provides practical insights for educators, institutional leaders, and policymakers to support the ethical, effective, and sustainable adoption of AI-enabled learning in higher education.
- 25 September 2026
- Published by Springer Nature in International Journal of Computational Intelligence Systems
Open AccessReimagining Education: A Comparative Study of Artificial Intelligence and Large Language Model-based Learning Systems and Traditional Education
This study presents a comparative analysis of Artificial Intelligence (AI) and Large Language Model (LLM)-based learning systems versus traditional education methods, focusing on comprehension, retention, personalization, and critical thinking. With AI and LLM technologies transforming instructional design, it is essential to examine their pedagogical effectiveness and implications for higher education. A convergent mixed-methods approach was employed, combining quantitative and qualitative data collected from 350 students and faculty across different Indian academic institutions. A structured survey assessed perceptions of learning effectiveness, engagement, and assessment accuracy across both systems. Findings indicate that nearly half of the respondents viewed AI/LLM-based systems as more effective in enhancing comprehension and retention, while a majority recognized their strength in personalization and adaptive feedback. Traditional methods, however, remained valuable for fostering mentorship, ethical awareness, and collaborative learning. The study highlights the complementary strengths of AI-driven and traditional approaches, suggesting that hybrid learning models can maximize educational outcomes by integrating technological adaptability with human-centred pedagogy. The study contributes empirical evidence on the comparative effectiveness of AI/LLM-based and traditional learning approaches and provides practical insights for educators, institutional leaders, and policymakers to support the ethical, effective, and sustainable adoption of AI-enabled learning in higher education.
- 14 August 2026
- Published by Taru Publications in Journal of Discrete Mathematical Sciences and Cryptography
- Vol. 29 (8) , 3149-3158
Encryption-assisted reversible data hiding using median prediction and grayscale invariance for high-security image systems
Ensuring data security and privacy is essential when transmitting sensitive visual information over insecure networks. Reversible Data Hiding (RDH) enables conf...
- 14 August 2026
- Published by Taru Publications in Journal of Discrete Mathematical Sciences and Cryptography
- Vol. 29 (8) , 3149-3158
Encryption-assisted reversible data hiding using median prediction and grayscale invariance for high-security image systems
Ensuring data security and privacy is essential when transmitting sensitive visual information over insecure networks. Reversible Data Hiding (RDH) enables conf...
- Open Access
- 11 August 2026
- Published by Springer Nature in International Journal of Networked and Distributed Computing
- Vol. 14 (2)
The Confluence of the Internet of Things and Medicare: Trends, Opportunities, and Future Research Pathways
Internet of Things (IoT) technologies and healthcare present revolutionary chances to improve operational efficiency, patient outcomes, and tailored medication transformation. This paper thoroughly investigates IoT in healthcare using Latent Dirichlet Allocation (LDA) to spot important trends and research gaps in current work. To achieve this, researchers have comprehensively analyzed 11,586 published papers from 2006 to 2024 which are extracted from Scopus database. Researchers have identified 2, 5, and 10 key topics to define significant areas of the research. Over time, it compares research topics to show how important areas, including wearable technology, artificial intelligence-powered analytics, blockchain for safe data management, and edge computing, have evolved. The paper additionally examines important issues, including data privacy issues, lack of interoperability, and restricted inclusiveness for underprivileged communities. Emphasizing inclusivity, ethical compliance, and pragmatic implementation tactics catered to different healthcare environments, a strategy framework is suggested to help solve these difficulties. This paper helps IoT implementation in healthcare advance by giving actionable insights, particular discoveries, and future research directions, thereby opening the path for more fair, efficient, and sustainable healthcare systems.
- 11 August 2026
- Published by Springer Nature in International Journal of Networked and Distributed Computing
- Vol. 14 (2)
Open AccessThe Confluence of the Internet of Things and Medicare: Trends, Opportunities, and Future Research Pathways
Internet of Things (IoT) technologies and healthcare present revolutionary chances to improve operational efficiency, patient outcomes, and tailored medication transformation. This paper thoroughly investigates IoT in healthcare using Latent Dirichlet Allocation (LDA) to spot important trends and research gaps in current work. To achieve this, researchers have comprehensively analyzed 11,586 published papers from 2006 to 2024 which are extracted from Scopus database. Researchers have identified 2, 5, and 10 key topics to define significant areas of the research. Over time, it compares research topics to show how important areas, including wearable technology, artificial intelligence-powered analytics, blockchain for safe data management, and edge computing, have evolved. The paper additionally examines important issues, including data privacy issues, lack of interoperability, and restricted inclusiveness for underprivileged communities. Emphasizing inclusivity, ethical compliance, and pragmatic implementation tactics catered to different healthcare environments, a strategy framework is suggested to help solve these difficulties. This paper helps IoT implementation in healthcare advance by giving actionable insights, particular discoveries, and future research directions, thereby opening the path for more fair, efficient, and sustainable healthcare systems.
- Open Access
- 10 August 2026
- Published by Springer Nature in Discover Internet of Things
- Vol. 6 (1)
A hybrid CNN-BiLSTM edge-cloud intrusion detection system with online incremental learning and SHAP explainability for smart city IoT
Smart city IoT deployments interconnect safety-critical infrastructure across millions of heterogeneous devices, creating an attack surface that signature-based intrusion detection systems (IDS) cannot defend against zero-day exploits, polymorphic malware, or concept drift. Prior deep-learning IDS proposals address the detection accuracy gap but typically evaluate on a single benchmark with a single random seed, omit structured ablation evidence, lack on-line adaptation, and provide no model-explanation interface for security analysts. This paper presents MI-IDS, a hybrid Convolutional Neural Network–Bidirectional Long Short-Term Memory (CNN-BiLSTM) ensemble deployed on a two-tier edge-cloud framework that integrates reservoir-sampling-based incremental learning and SHAP explainability under a single experimentally validated pipeline. The experimental results reveals that across five random seeds, MI-IDS achieves 96.7 ± 0.2% accuracy and 95.8 ± 0.2% F1-score on a 73,100-instance composite benchmark spanning seven traffic classes, one of which is a held-out group of synthetically mutated attack variants used as a partial-novelty proxy rather than a genuine zero-day family. A held-out UNSW-NB15 partition ( n = 82,332) yields 97.4 ± 0.2% accuracy; because UNSW-NB15 also contributes to the composite, this figure reflects within-benchmark held-out performance rather than independent cross-dataset generalisation. A six-variant ablation study isolates the contribution of each architectural component. Under a 48-hour concept-drift simulation, reservoir sampling bounds accuracy loss to 1.2% points versus a 20.4-point degradation for the non-adaptive baseline. The hybrid edge-cloud deployment achieves 14.1 ms mean detection latency and 61.2% lower bandwidth than cloud-only deployment, and SHAP attributions lowered analyst mean time-to-decision by 57.1% in a small preliminary study with five experts, a result we treat as indicative rather than confirmatory. All improvements over six baselines are statistically significant at Bonferroni-corrected α = 0.0083. A central contribution is the integration and disciplined evaluation of multi-seed validation, structured ablation, on-line incremental learning, and model-agnostic explainability within a single smart city IoT pipeline, a combination that remains uncommon in the prior IDS literature rather than one we can demonstrate to be unprecedented.
- 10 August 2026
- Published by Springer Nature in Discover Internet of Things
- Vol. 6 (1)
Open AccessA hybrid CNN-BiLSTM edge-cloud intrusion detection system with online incremental learning and SHAP explainability for smart city IoT
Smart city IoT deployments interconnect safety-critical infrastructure across millions of heterogeneous devices, creating an attack surface that signature-based intrusion detection systems (IDS) cannot defend against zero-day exploits, polymorphic malware, or concept drift. Prior deep-learning IDS proposals address the detection accuracy gap but typically evaluate on a single benchmark with a single random seed, omit structured ablation evidence, lack on-line adaptation, and provide no model-explanation interface for security analysts. This paper presents MI-IDS, a hybrid Convolutional Neural Network–Bidirectional Long Short-Term Memory (CNN-BiLSTM) ensemble deployed on a two-tier edge-cloud framework that integrates reservoir-sampling-based incremental learning and SHAP explainability under a single experimentally validated pipeline. The experimental results reveals that across five random seeds, MI-IDS achieves 96.7 ± 0.2% accuracy and 95.8 ± 0.2% F1-score on a 73,100-instance composite benchmark spanning seven traffic classes, one of which is a held-out group of synthetically mutated attack variants used as a partial-novelty proxy rather than a genuine zero-day family. A held-out UNSW-NB15 partition ( n = 82,332) yields 97.4 ± 0.2% accuracy; because UNSW-NB15 also contributes to the composite, this figure reflects within-benchmark held-out performance rather than independent cross-dataset generalisation. A six-variant ablation study isolates the contribution of each architectural component. Under a 48-hour concept-drift simulation, reservoir sampling bounds accuracy loss to 1.2% points versus a 20.4-point degradation for the non-adaptive baseline. The hybrid edge-cloud deployment achieves 14.1 ms mean detection latency and 61.2% lower bandwidth than cloud-only deployment, and SHAP attributions lowered analyst mean time-to-decision by 57.1% in a small preliminary study with five experts, a result we treat as indicative rather than confirmatory. All improvements over six baselines are statistically significant at Bonferroni-corrected α = 0.0083. A central contribution is the integration and disciplined evaluation of multi-seed validation, structured ablation, on-line incremental learning, and model-agnostic explainability within a single smart city IoT pipeline, a combination that remains uncommon in the prior IDS literature rather than one we can demonstrate to be unprecedented.
- Open Access
- 21 June 2026
- Published by Springer Nature in Discover Artificial Intelligence
- Vol. 6 (1)
A lightweight transformer based language model inspired by the Qwen architecture for resource constrained environments
Large Language Models have proved to be extremely effective in the realm of NLP; however, they are both computationally expensive and highly complex in terms of architecture, thus limiting their usage in systematic analysis and research due to a lack of computational resources. In this work, we present MinimalLLM, a small-scale decoder-only language model built on a transformer architecture, similar to that used by the Qwen model family. The model attempts to provide a balance between modern architectural quality and efficiency and transparency, with an emphasis being placed on experiment control and interpretability. Our implementation features Grouped-Query Attention (GQA), RoPE, SwiGLU, and RMSNorm. A novel hybrid optimisation approach based on the Muon method used for structured 2D weight matrices and AdamW for all other parameters is proposed to facilitate stable convergence in computational budget-constrained environments. MinimalLLM is pre-trained on a streaming subset of the 500,000 tokens of the Cosmopedia-v2 synthetic corpus through mixed precision training and gradient accumulation on an NVIDIA Tesla T4 GPU. Although it is trained with just 32.15 million parameters, MinimalLLM obtains a validation loss of 0.0945 and perplexity of 1.10 on the domain of Cosmopedia-v2. These figures must be taken as indicative of the optimal convergence achieved by the network for its specific training conditions due to its over-parametrisation, the Zipfian sparsity of the corpus, and the nature of its constrained synthetic domain. Analysis of attentional patterns and of gradients, and confidence calibration, support that optimisation was maintained and that token-level representation has analytic meaning. We show that state-of-the-art Transformer architecture design principles can be implemented effectively in a small-scale setting and describe a reproducible approach to language modeling for research and educational purposes. Additional analysis such as comparison with baselines, performance evaluation using statistical methods, tests with other datasets, and ablation studies reinforce the robustness of our approach.
- 21 June 2026
- Published by Springer Nature in Discover Artificial Intelligence
- Vol. 6 (1)
Open AccessA lightweight transformer based language model inspired by the Qwen architecture for resource constrained environments
Large Language Models have proved to be extremely effective in the realm of NLP; however, they are both computationally expensive and highly complex in terms of architecture, thus limiting their usage in systematic analysis and research due to a lack of computational resources. In this work, we present MinimalLLM, a small-scale decoder-only language model built on a transformer architecture, similar to that used by the Qwen model family. The model attempts to provide a balance between modern architectural quality and efficiency and transparency, with an emphasis being placed on experiment control and interpretability. Our implementation features Grouped-Query Attention (GQA), RoPE, SwiGLU, and RMSNorm. A novel hybrid optimisation approach based on the Muon method used for structured 2D weight matrices and AdamW for all other parameters is proposed to facilitate stable convergence in computational budget-constrained environments. MinimalLLM is pre-trained on a streaming subset of the 500,000 tokens of the Cosmopedia-v2 synthetic corpus through mixed precision training and gradient accumulation on an NVIDIA Tesla T4 GPU. Although it is trained with just 32.15 million parameters, MinimalLLM obtains a validation loss of 0.0945 and perplexity of 1.10 on the domain of Cosmopedia-v2. These figures must be taken as indicative of the optimal convergence achieved by the network for its specific training conditions due to its over-parametrisation, the Zipfian sparsity of the corpus, and the nature of its constrained synthetic domain. Analysis of attentional patterns and of gradients, and confidence calibration, support that optimisation was maintained and that token-level representation has analytic meaning. We show that state-of-the-art Transformer architecture design principles can be implemented effectively in a small-scale setting and describe a reproducible approach to language modeling for research and educational purposes. Additional analysis such as comparison with baselines, performance evaluation using statistical methods, tests with other datasets, and ablation studies reinforce the robustness of our approach.
- Open Access
- 11 December 2025
- Published by Springer Nature in Discover Artificial Intelligence
- Vol. 5 (1)
Deep vision against deception using CNN strategies for fake social media profile detection
- 11 December 2025
- Published by Springer Nature in Discover Artificial Intelligence
- Vol. 5 (1)
Open AccessDeep vision against deception using CNN strategies for fake social media profile detection