Published December 10, 2024 | Version v1
Dataset Open

Vector Maps (used to create jigsaw puzzles)

  • 1. ROR icon Ca' Foscari University of Venice

Description

These maps have been created using cadmapper, and they are used to test the jigsaw puzzle solver algorithm on real data.

The jigsaw puzzle solver approach is described in the paper:

Nash Meets Wertheimer: Using Good Continuation in Jigsaw Puzzles, Khoroshiltseva et al., ACCV 2024. 

The data here are 13 maps (in .dxf format) and the relative rendered images. In the paper the vector data was extracted from the .dxf file and the images used for color-based puzzle solving are the ones provided here. 

As part of the original data comes from Openstreetmaps, we release it the Open Database License (in the License.txt the other sources are also mentioned, generated from cadmapper).

Files

images.zip

Files (2.5 MB)

Name Size Download all
md5:4dd5f0bdf3c768562ab0052e038030a7
120.0 kB Preview Download
md5:07fd09176678b5f4a859c8a50dad6227
382 Bytes Preview Download
md5:fb9c09858123b115d99230d250ab0ced
2.4 MB Preview Download

Additional details

Related works

Is part of
Conference proceeding: arXiv:2410.16857 (arXiv)

Funding

European Commission
RePAIR - Reconstructing the Past: Artificial Intelligence and Robotics Meet Cultural Heritage 964854

Software

Repository URL
https://github.com/RePAIRProject/RL_puzzle_solver
Programming language
Python

References

  • @InProceedings{Khoroshiltseva_2024_ACCV, author = {Khoroshiltseva, Marina and Palmieri, Luca and Aslan, Sinem and Vascon, Sebastiano and Pelillo, Marcello}, title = {Nash Meets Wertheimer: Using Good Continuation in Jigsaw Puzzles}, booktitle = {Proceedings of the Asian Conference on Computer Vision (ACCV)}, month = {December}, year = {2024}, pages = {3223-3238} }