g
ghafoorkhan251

Ghafoor Khan

@ghafoorkhan251

AI and Computer Vision Engineer , Building Smarter, Safer, and Faster

Pakistan
Engels, Urdu, Pasjtoe
Sommige informatie wordt in het Engels weergegeven.
Over mij
I am an AI Engineer with a strong background in computer vision and deep learning. My expertise lies in developing and optimizing models for object detection, image segmentation, and classification tasks using modern architectures and frameworks such as YOLO, OpenCV, and PyTorch. I have hands-on experience in building real-time AI systems, fine-tuning models, and deploying them for practical applications. Passionate about innovation, I continuously explore emerging AI technologies to create intelligent and efficient solutions that drive automation and enhance decision-making.... Lees meer

Skills

g
ghafoorkhan251
Ghafoor Khan
offline • 
Gemiddelde reactietijd: 2 uur

Bekijk mijn diensten

AI-integraties
I will do computer vision, object detection, classification and segmentation for you

Werkervaring

Research Assistant

NRCIA • Fulltime

Aug 2025 - Present9 mos

Developed and deployed multiple AI-powered computer vision systems for real-world applications including rebar counting, agriculture detection, and employee monitoring Built and optimized YOLO-based detection and segmentation models achieving 90%+ accuracy in industrial scenarios Designed end-to-end pipelines (data collection → training → deployment) for scalable AI solutions Successfully deployed models on edge devices (Jetson Nano, OAK-D, Raspberry Pi) enabling real-time, low-latency performance Improved system efficiency through model optimization, dataset engineering, and inference tuning 🏆 Key Achievements 🚀 Delivered 5+ production-level AI systems used in industrial and surveillance environments 📊 Achieved high-accuracy detection (90%+) in complex real-world conditions (dense objects like rebar bundles) ⚡ Enabled real-time edge AI deployment, reducing dependency on cloud and improving response time 🧠 Built solutions combining object detection + VLM (Vision Language Models) for intelligent automation 🔧 Specialized in high-density object counting problems (e.g., 500+ rebars per image with precise detection)