yolo26s-highway-heavy-vehicle-truck-tr
heavy transport and truck road detection
1 min read
An object-detection model trained on a large volume of road data to detect and classify trucks and heavy transport across 4 classes. The architecture weighs 19.3 MB, enabling efficient processing of streaming video and security camera footage. The product is built on the ultralytics framework (a library for working with YOLO models) and optimized for a 512×512 pixel resolution. This solution is useful for logistics systems, smart roads, and traffic flow analysis.
Key features
- High accuracy: mAP@50 reaches 93.78% on the validation dataset.
- Truck specialization: 4 detailed classes for different types of heavy vehicles.
- Compact size: the model is only 19.3 MB for fast inference processing.
- Clean training process: trained on 9,900 images featuring over 117,000 labeled objects.
What you can do
- Run streaming video analysis from traffic cameras to count and classify traffic.
- Integrate the model into logistics systems for monitoring truck access at terminals.
- Monitor road corridors and collect statistics on road congestion.
Example
from ultralytics import YOLO
model = YOLO("yolo26s-highway-heavy-vehicle-truck-tr.pt")
results = model.predict(
source="ornek_gorsel.jpg",
conf=0.25,
iou=0.45,
imgsz=640,
)
results[0].show()Read more
Details• mAP@50: 93.78%
• mAP@50-95: 67.33%
• F1 Score: 88.97%
• Training: 50 epochs with a batch size of 32 on 9,900 images










Comments