EMBEDDED COMPUTER VISION
Embedded systems, image processing, and real-time vision on resource-constrained hardware.
About me
My work has evolved from embedded image processing under resource-constrained computing to production computer vision, edge AI, and explainable systems designed for real-world deployment.
Embedded systems, image processing, and real-time vision on resource-constrained hardware.
Multi-camera detection, tracking, segmentation, recognition, and optimized AI deployment.
Explainable AI for multimodal models, system health monitoring, and prognostics.
Explainable and reliable computer vision solutions for edge deployment.
Who I am
My experience covers the full computer vision workflow, from research and model development to optimization, deployment, and explainability, with an emphasis on building production-ready systems.
I approach computer vision as a complete system rather than only a model. That means considering data quality, preprocessing, model behavior, hardware constraints, inference efficiency, and integration from the beginning of development.
Working across both research and production has shaped how I evaluate solutions, considering not only accuracy, but also reliability and the trade-offs between accuracy, inference speed, memory usage, and energy consumption.
Data
Preprocessing
Model
Hardware
Integration
Accuracy
Reliability
Inference speed
Memory usage
Energy consumption
Experimentation
Failure analysis
Optimization
Deployment
Iteration