Irfan H.

I'm Irfan Hussain, a Computer Vision Engineer & DevOps enthusiast with hands-on experience designing real-time AI systems for video analytics, satellite imagery, and drone-based object detection and tracking. Anomaly action detection and recognition in retail stores USA with upto 64 cameras inferencing in real-time. I’ve built and deployed high-performance computer vision solutions that are both scalable and production-ready. Here’s what I bring to the table: Computer Vision & Deep Learning • Real-time object detection and action recognition (YOLO, YOWOv3, SlowFast, Kinetics-13D) • Satellite image analysis and small object detection using customized models with patching techniques. • Drone/UAV tracking systems using YOLO-TLP(MS-thesis) and advanced multi-object tracking • Data harmonization with SinGAN and 3D synthetic dataset generation DevOps & Deployment • Real-time camera streaming with OpenCV, CUDA, multi-threading • Scalable backend systems using AWS RDS, Python, and REST APIs • Integration with cloud services like Facebook APIs and AWS databases Research & Prototyping • Worked on cutting-edge R&D at NASTP and GIKI in aerospace and robotics vision • Designed and tested systems on real-world data including drone footage & CCTV • Visual analytics and model performance tuning for high mAP & FPS Tech Stack: Python, OpenCV, YOLO, PyTorch, TensorFlow, CUDA, AWS, FFmpeg, VLC, REST APIs, NetCDF I combine deep learning expertise with system-level engineering to deliver intelligent computer vision applications that actually work in the wild. If you’re building something in AI, vision, or video — let’s connect and make it real.