Bilal Z.

I build Computer Vision systems that work with real cameras, video streams, and challenging image data not just models running in notebooks. My completed projects include real-time video analytics, YOLO-based object detection, image registration and reconstruction, and AWS-based face recognition and emotion analysis. I work with Python, OpenCV, PyTorch, YOLO, object tracking, and cloud/edge deployment. Top Rated Plus | 100% Job Success | $20K+ Earned | 1,800+ Upwork Hours SELECTED PROJECT EXPERIENCE 1. Real-Time CCTV, Video Analytics & Object Detection Developed computer vision applications for surveillance and live-video processing, including object detection, camera integration, tracking, event detection, and video analysis. My work also includes a streaming and live-TV advertisement detection system that identifies commercial segments and transitions in video feeds. 2. Custom YOLO Object Detection & Tracking Worked on custom YOLO detection systems, dataset preparation, model training, and inference pipelines. Applications include security surveillance and construction-site PPE monitoring. I focus on practical problems such as false detections, occlusion, camera angles, lighting variation, and inference performance. 3. Image Registration & Reconstruction Completed an image-reconstruction project involving tiled engineering images with unknown ordering, rotation, noise, and alignment differences. The work involved image matching, reconstruction, cropping, and visual consistency without depending solely on image metadata. 4. AWS Video Processing, Face Recognition & Emotion Analysis Developed webcam and video-processing functionality involving face recognition, emotion analysis, known/unknown face handling, detection history, APIs, and AWS services. WHAT I CAN HELP YOU BUILD - YOLO & Object Detection: Custom models, training, fine-tuning, classification, segmentation, object counting, and evaluation. - CCTV / RTSP / Video Analytics: IP cameras, webcam processing, multi-object tracking, ByteTrack, DeepSORT, zone monitoring, event detection, and automated alerts. - OCR & Document Processing: PaddleOCR, Tesseract, OpenCV preprocessing, document layout analysis, text extraction, and structured JSON output. - Image Processing & Computer Vision: Image registration, stitching, feature matching, homography, camera calibration, perspective correction, and reconstruction. - AI Deployment: PyTorch, TensorFlow, ONNX Runtime, TensorRT, FastAPI, Flask, Docker, AWS, and GPU inference optimization. HOW I WORK I focus on delivering working applications, reproducible code, trained models, evaluation results, and integrations not just technical recommendations. I start by understanding your real input data, target outputs, and deployment requirements. Then I build and validate the solution against representative images or video before expanding it. Have a Computer Vision project? Send me a sample image, video, RTSP camera requirement, or dataset, along with the result you want to achieve. I'll help identify a practical implementation approach and clear first milestone.