metadata

license: cc-by-4.0

Probabilistic Road Classification in Historical Maps Using Synthetic Data and Deep Learning

Dominik J. Mühlematter, Sebastian Schweizer, Chenjing Jiao, Xue Xia, Magnus Heitzler, Lorenz Hurni - 2024

[Paper on ArXiv]

In correspondence with the code we released on GitHub, the usage of the models within our pipeline is described in the repository. Please note that this repository contains only the models for our final results, not for all intermediate results.

Pretraining

This folder contains the pretrained models, including:

ResNet18 Classification Backbone: Pretrained on ImageNet (For more details, see the PyTorch GitHub repository) .
Binary Road Segmentation Model: Initialized with the ImageNet classification backbone and trained using cascaded training with Swiss Map data.

Binary_road_segmentation

This folder contains the final model weights used for extracting roads from the Siegfried Map.

Road_classification_ensemble

This folder contains all the model weights for the final road classification ensemble trained on the Siegfried Map.

Citation

If you find our work useful or interesting, or if you use our code, please cite our paper as follows:

@misc{muhlematter2024probabilistic,  
  title   = {Probabilistic road classification in historical maps using synthetic data and deep learning},  
  author  = {Dominik J. Mühlematter, Sebastian Schweizer, Chenjing Jiao, Xue Xia, Magnus Heitzler, Lorenz Hurni},  
  year    = {2024},  
  note    = {arXiv:xxxx.xxxxx}  
}