Object Detection
ultralytics
English
driver-monitoring
drowsiness-detection
yolo
computer-vision
safety
Instructions to use raj5517/safedrive-model with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- ultralytics
How to use raj5517/safedrive-model with ultralytics:
from huggingface_hub import hf_hub_download from ultralytics import YOLO # pick the weights file from this repo's "Files and versions" tab weights = hf_hub_download("raj5517/safedrive-model", "<weights>.pt") model = YOLO(weights) source = 'http://images.cocodataset.org/val2017/000000039769.jpg' model.predict(source=source, save=True) - Notebooks
- Google Colab
- Kaggle
SafeDrive Model Suite
Models powering the safedrive-ai Python SDK for real-time driver monitoring.
Files
| File | Purpose | Accuracy |
|---|---|---|
yolo_safedrive.pt |
YOLOv8-nano, 9-class detection | mAP50=0.940 |
mobilenet_webcam.pth |
MobileNetV3 eye classifier (webcam fine-tuned) | 97.99% |
mobilenet_best.pth |
MobileNetV3 eye classifier (lab-trained) | 97.99% |
drowsiness_cnn_best.pth |
Custom CNN eye classifier | 96.74% |
face_landmarker.task |
MediaPipe face landmark model | — |
Classes (YOLO)
eye_open, eye_half, eye_closed, mouth_open, mouth_closed, phone, cigarette, seatbelt_on, seatbelt_off
Usage
pip install safedrive-ai
- Downloads last month
- 66