You can deploy this model and start earning money today!

akahukas/
yolo26n-seg-rock-positioning

Name: yolo26n-seg-rock-positioning
Availability: InStock

akahukas/rock-segmentation-dataset

Ultralytics/YOLO26

quantized

agpl-3.0

Views

No views yet

YOLO26n-seg for Rock Segmentation

Original RGB	Segmentation Mask	Contour Overlay

This model is a fine-tuned YOLO26n-seg developed as part of a Master's Thesis titled:
Evaluating Rock Position Estimation Pipeline Using 3D Sensors for Excavation Task

The model has been used to segment rocks by using images captured during the rock-picking task of a real excavator as an input. You can find further information and see the model in action at the Thesis Project Page

Training Process

Data

The model has been finetuned with the following dataset: rock-segmentation-dataset

Hardware & Software

The model has been trained on a RTX 3080 10GB GPU, utilizing the Ultralytics framework.

Acknowledgements

The thesis work has been completed in the Autonomous Mobile Machines Group, at Tampere University.
The thesis work has been funded by the Horizon Europe Project XSCAVE under Grant 101189836.
Thanks to Novatron Oy for providing access to the excavator and the data interface.

Citation

If you use this model, please cite the original thesis as the primary research source and this repository for the model artifact.

bibtex

@mastersthesis{hakamaki2026thesis,
  author    = {Saku Hakamäki},
  title     = {Evaluating Rock Position Estimation Pipeline Using 3D Sensors for Excavation Task},
  school    = {Tampere University},
  year      = {2026},
  url       = {https://urn.fi/URN:NBN:fi:tuni-202604274316},
}

@software{hakamaki2026model,
  author       = {Saku Hakamäki},
  title        = {YOLO26n-seg for Rock Segmentation},
  year         = {2026},
  publisher    = {Hugging Face},
  version      = {1.0.0},
  url          = {https://huggingface.co/akahukas/yolo26n-seg-rock-positioning}
}