object detection and overlays on video streams
Find a file
2026-07-07 11:44:22 +00:00
object_detection object_detection/RT_10FPSblacks_yolo5n.py 2026-03-11 16:09:24 +05:30
README.md Update README.md 2026-07-07 11:44:22 +00:00

Pipe an arbitrary youtube livestream to an html page (yt-dlp and ffmpeg for any changes/ extraction)

detect objects in it based on yolo v5, 8 or 9 image detection models. Do various things like detect from a list, from a NOT list, save images with bbox to object-wise folder, black out objects in the stream, etc.

To do: fps tests overlays tests