Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a real time face mask detection system using yolov8
Haseeb Qaiser
About this gig
Hello! Are you looking for a production-ready, real-time face mask detection system with a modern web interface?
I will build a complete real-time face mask detection system powered by YOLOv8 + ONNX running on CPU no GPU required! The system streams live webcam video through a FastAPI backend via WebSocket and displays annotated results in a clean HTML/CSS frontend all in your browser.
This is based on a fully working project with a complete end-to-end pipeline.
What You Will Get
YOLOv8 Nano model trained for mask / no_mask detection
Model exported to ONNX for optimized CPU inference
FastAPI backend server (app.py)
WebSocket real-time frame streaming
OpenCV webcam capture & frame processing
HTML/CSS frontend with live bounding box overlay
Full annotated source code (clean & commented)
requirements.txt for one-command install
Setup & run instructions (README)
Real-Time Pipeline
Webcam YOLO ONNX Model FastAPI WebSocket HTML Frontend
Tech Stack
YOLOv8 (Ultralytics) Object Detection
ONNX CPU-Optimized Inference
FastAPI Backend API Server
WebSocket Real-Time Frame Streaming
OpenCV (cv2) Webcam Capture
HTML / CSS Frontend UI
Python 3.x Core Language
Get to know Haseeb Qaiser
Haseeb Qaiser
AI Engineer
- FromPakistan
- Member sinceApr 2022
Languages
English, Pashto, Urdu
👋 Welcome! I am an AI Engineer with hands-on experience in Machine Learning, Deep Learning, and Computer Vision, focused on building practical and reliable AI solutions. I have worked on projects such as deepfake face and video detection, image classification, object detection, AI-powered chatbots, and data annotation for computer vision datasets. I am experienced in preparing, cleaning, labeling, and annotating image and video data to ensure high-quality training datasets. I primarily use Python along with frameworks like TensorFlow, PyTorch, OpenCV, and Scikit-learn to develop, train.
My Portfolio
FAQ
Q: Does this need a GPU to run?
Ans: No! The model is exported to ONNX format which runs efficiently on CPU. No CUDA or GPU required — it works on any modern laptop
Q: What does the frontend look like?
Ans: A clean browser page that shows your live webcam feed with real-time bounding boxes and "mask" / "no mask" labels — accessible at http://127.0.0.1:8000.
Q: Can I retrain the model on my own data?
Ans: Yes, with the Premium package I'll retrain YOLOv8 on your custom labeled dataset and re-export the updated ONNX model.
Q: Is the full source code included?
Ans: Yes, all packages include app.py, best.onnx, index.html, style.css, and requirements.txt — the complete project, nothing hidden
Q: Can this be deployed to a server or cloud?
Ans: Yes. Docker deployment and cloud hosting are available in the Premium package. It's a future improvement already planned in the project roadmap.
