Glowing YOLO

Lightweight demo that runs Ultralytics YOLOv8 on a video and renders label-free, per-class colored bounding boxes with a soft neon glow. Includes sample weights, a test clip, the rendered output, and a preview GIF.

Preview

Quick start

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python run_glowing_yolo.py --input myvideo.mp4 --output runs/glowing_yolo/output.mp4 --model yolov8n.pt --conf 0.25 --glow-radius 12 --line-thickness 2

How it works

• Loads a YOLOv8 model (default yolov8n.pt), runs per-frame inference on the input video.
• Deterministic per-class colors (overrides for common traffic classes, otherwise hash-based hues).
• Renders two-layer boxes: a blurred glow plus a crisp outline on top.
• Writes the processed video to runs/glowing_yolo/output.mp4; creates parent directories when missing.

Assets included

• myvideo.mp4 – sample input clip.
• yolov8n.pt and yolov8l.pt – model weights (use --model to switch).
• runs/glowing_yolo/output.mp4 – rendered output from the sample clip.
• runs/glowing_yolo/preview.gif – 4-second excerpt (at 4s–8s) for quick viewing.

Make your own preview GIF

ffmpeg -y -ss 4 -t 4 -i runs/glowing_yolo/output.mp4 -vf "fps=15,scale=640:-1:flags=lanczos,palettegen" /tmp/palette.png
ffmpeg -y -ss 4 -t 4 -i runs/glowing_yolo/output.mp4 -i /tmp/palette.png -lavfi "fps=15,scale=640:-1:flags=lanczos [x]; [x][1:v] paletteuse" runs/glowing_yolo/preview.gif