Quantum algorithms, NISQ-era circuits, and variational quantum classifiers. Quantum key distribution via the BB84 protocol and quantum-secured hybrid communication systems.

_{Quantum ML · VQC · BB84 · PennyLane · Qiskit}

Attack and defense techniques for reliable, safe AI systems — adversarial robustness, federated learning security, post-hoc defense, and fault injection.

_{Adversarial Attack · FL Security · Fault Injection · Trustworthy AI}

Autonomous driving systems built on Federated Learning and Large Language Models. Optimizing FL across heterogeneous and homogeneous models in diverse driving environments.

_{Federated Learning · LLM · Autonomous Driving · Simulation}

Design and implementation of real-time AI systems on FPGA, AI accelerators, and neuromorphic computing architectures. Hardware-optimized voice risk detection systems.

_{FPGA · AI Accelerator · Neuromorphic · CUDA}

AI in healthcare — pneumonia classification from chest X-rays, federated learning for medical imaging, and hybrid quantum-classical classification with privacy preservation.

_{Medical Imaging · Chest X-ray · Federated Learning · Privacy}

Adversarial robustness analysis of Automatic Modulation Classification in wireless communication. Quantum-secured communication systems for tactical military networks.

_{AMC · Wireless Security · Tactical Network · Quantum Security}

Visual understanding — object detection, semantic and instance segmentation, 3D reconstruction, and generative models for image synthesis and style transfer.

_{Object Detection · Segmentation · 3D Vision · Image Generation}

Intelligent robotic systems — motion planning, dexterous manipulation, sim-to-real transfer learning, and multi-agent coordination for collaborative tasks.

_{Motion Planning · Manipulation · Sim-to-Real · Multi-Agent}

LLM capabilities — prompt engineering, fine-tuning, retrieval-augmented generation, multi-modal LLMs, and alignment for safe and reliable AI systems.

_{Prompt Engineering · Fine-tuning · RAG · Multi-modal · Alignment}

AI/ML compilers (TVM / XLA / MLIR-style graph optimization), hardware-aware compilation for FPGA, GPU, and NPU backends, quantum circuit compilation for NISQ devices, and classical compiler and PL theory on LLVM.

_{AI/ML Compiler · Hardware-aware · Quantum Compiler · LLVM / PL}