Research
Complete record of journal and conference work in explainable computer vision, predictive maintenance, federated and edge-cloud AI, multi-agent systems, and intelligent transportation. Publication status is shown on every entry.
An explainable deepfake-detection framework using vision-transformer representations and attention-based interpretation to improve transparency in manipulated-media classification.
Contribution: Model development, explainability pipeline, and evaluation
An IoT-oriented small-object segmentation framework that combines deterministic data auditing, modality-specific patch embeddings, a dual-stream shared-weight Vision Transformer backbone, and quality-aware edge-guided bidirectional cross-attention for robust multimodal perception under sensor corruption and deployment constraints.
Contribution: Model architecture, cross-attention fusion design, and edge-deployment evaluation
A temporal-convolutional approach for identifying early fault patterns in industrial-machinery sensor data.
Contribution: Model development, experimental design, and evaluation
An evaluation of routing stability, task coordination, and communication behavior in swarm-based multi-agent dialogue systems.
Contribution: Experimental design, evaluation, and result analysis
A unified vehicle-mounted sensing and edge-computing pipeline that integrates visual, inertial and positioning streams, couples reliability-aware alignment and fusion with communication-efficient reporting, and delivers robust road-anomaly monitoring for consumer vehicles.
Contribution: System architecture, multi-modal fusion design, and edge-deployment evaluation
A systematic survey of dynamic token computation methods in transformers, organizing approaches by taxonomy, analyzing training stability, and evaluating budget-aware inference trade-offs.
Contribution: Taxonomy, stability analysis, and budget-aware evaluation
An explainable deepfake-detection framework using vision-transformer representations and attention-based interpretation to improve transparency in manipulated-media classification.
Contribution: Model development, explainability pipeline, and evaluation
An IoT-oriented small-object segmentation framework that combines deterministic data auditing, modality-specific patch embeddings, a dual-stream shared-weight Vision Transformer backbone, and quality-aware edge-guided bidirectional cross-attention for robust multimodal perception under sensor corruption and deployment constraints.
Contribution: Model architecture, cross-attention fusion design, and edge-deployment evaluation
A predictive-maintenance approach combining feature-selection methods and imbalance-aware learning for early engine-failure prediction on large-scale sensor data.
Contribution: Feature engineering, model development, and result analysis
A temporal-convolutional approach for identifying early fault patterns in industrial-machinery sensor data.
Contribution: Model development, experimental design, and evaluation
An edge-cloud federated-learning framework designed to support collaborative maritime intelligence while protecting distributed data.
Contribution: System architecture, experimental design, and evaluation
An evaluation of routing stability, task coordination, and communication behavior in swarm-based multi-agent dialogue systems.
Contribution: Experimental design, evaluation, and result analysis
A unified vehicle-mounted sensing and edge-computing pipeline that integrates visual, inertial and positioning streams, couples reliability-aware alignment and fusion with communication-efficient reporting, and delivers robust road-anomaly monitoring for consumer vehicles.
Contribution: System architecture, multi-modal fusion design, and edge-deployment evaluation
A systematic survey of dynamic token computation methods in transformers, organizing approaches by taxonomy, analyzing training stability, and evaluating budget-aware inference trade-offs.
Contribution: Taxonomy, stability analysis, and budget-aware evaluation
A semantic-aware control framework for UAV-assisted IRS networks that improves cooperative V2X perception by leveraging label-derived context for resource-efficient communication.
Contribution: System design, control strategy, and evaluation
An explainable congestion-management approach for VANETs that combines neuro-symbolic learning with TabNet to deliver transparent, interpretable decisions for network traffic control.
Contribution: Model development, explainability design, and evaluation
20 publications listed.