← Back to projects
Computer visionIndependent founder project
SmartCam
Live object detection in the browser: YOLOv8 + FastAPI + OpenCV streams annotated frames.
- Problem
- Detecting what is in a live video stream needs separate software and setup.
- Founder’s role
- Dmitri built a YOLOv8 + FastAPI + OpenCV service that streams annotated frames to the browser.
- Result
- Objects are detected live right in the browser, with no separate app.
Problem
Many detection demos are offline clips or heavy client-side models. A useful demo needs a live stream, readable overlays and CPU-aware cadence.
Solution
A server-side YOLOv8 nano loop resizes the frame, detects, draws boxes and labels, and emits MJPEG so the browser only displays the annotated stream.
Result
A browser-visible live demo that stays stable on an imperfect stream and throttles work when the camera is faster than the CPU.
Technologies
- Python
- FastAPI
- YOLOv8
- OpenCV
- MJPEG
- Docker