Fire Smoke Yolov8
A YOLOv8n detector finetuned for early domestic fire and smoke: cooking, electrical, and cigarette scenarios. Part of the Model Garden collection of finetuned computer-vision models.
Fonte do modelo
- Tipo
- Other
- Tamanho do arquivo
- 5,97 MB
- Versão
- vRevision f1c6426
- Atualizado
- 10 de out. de 2026
- Downloads
- 328
- Fonte
- Hugging Face
Descrição da fonte
Early Fire and Smoke (YOLOv8)
A YOLOv8n detector finetuned for early domestic fire and smoke: cooking, electrical, and cigarette scenarios. Part of the Model Garden collection of finetuned computer-vision models.
Classes
fire, smoke
Intended use
Designed to pair with a temporal state machine that escalates through clear, smoke present, early warning (growing smoke), fire confirmed, so it alerts on persistent, growing smoke rather than a single noisy frame.
Training
- Base: YOLOv8n
- Data: D-Fire + Domestic Fire and Smoke
- Runtime: ~70 FPS on Apple Silicon (MPS)
Usage
from ultralytics import YOLO
model = YOLO("fire_smoke_yolov8.pt")
results = model("your_video.mp4", stream=True)
for r in results:
r.plot()
Limitations
Vision alone is never a life-safety system. This is a research tool, not a substitute for a certified smoke or fire detector.
Fontes
1 fonte
Fontes
1 fonteVerificado 10 de out.
Artefatos de modelo
1 artefato
Artefatos de modelo
1 artefatoTrechos de fonte
3 trechos
Trechos de fonte
3 trechos--- license: agpl-3.0 tags: - ultralytics - yolov8 - object-detection - fire-detection - smoke-detection library_name: ultralytics pipeline_tag: object-detection --- # Early Fire and Smoke (YOLOv8) A YOLOv8n detector finetuned for **early domestic fire and smoke**: cooking, electrical, and cigarette scenarios. Part of the [Model Garden](https://github.com/computer-vision-with-marco) collection of finetuned computer-vision models. ## Classes `fire`, `smoke` ## Intended use Designed to pair with a temporal state machine that escalates through **clear, smoke present, early warning (growing smoke), fire confirmed**, so it alerts on persistent, growing smoke rather than a single noisy frame. ## Training - **Base:** YOLOv8n - **Data:** D-Fire + Domestic Fire and Smoke - **Runtime:** ~70 FPS on Apple Silicon (MPS) ## Usage ```python from ultralytics import YOLO model = YOLO("fire_smoke_yolov8.pt") results = model("your_video.mp4", stream=True) for r in results: r.plot() ``` ## Limitations Vision alone is never a life-safety system. This is a research tool, **not** a substitute for a certified smoke or fire detector.
Source context: 328 downloads · 1 likes · Pipeline object-detection · Library ultralytics · Repo mfranzon/fire-smoke-yolov8
Source context: 132 downloads · 1 likes · Pipeline object-detection · Library ultralytics · Repo mfranzon/fire-smoke-yolov8
