Research Assistant
Edge Intelligent System Lab, Oakland University
Efficient AI
Model Optimization
LLM Systems
Quantization
Focus
Efficient AI and LLM systems for resource-constrained environments, focusing on
model compression, adaptive retrieval, and low-latency inference for edge and
domain-specific applications.
- Efficient Deep Learning: Develops and evaluates post-training quantization and model-compression strategies for CNN–Transformer hybrid and Vision Transformer models, emphasizing latency–accuracy trade-offs and deployment on edge devices.
- Efficient LLM and RAG Systems: Investigates adaptive retrieval, evidence sufficiency, dynamic chunk selection, and efficient LLM serving to reduce retrieval and generation costs while maintaining answer quality in domain-specific applications.
- Edge AI Applications: Builds efficient AI systems for connected health and autonomous driving, including knowledge-driven activity assessment and lightweight, real-time perception systems.