/Jigsaw-Solver

Solving Jigsaw puzzle using Neural Nets

Primary LanguageJupyter NotebookMIT LicenseMIT

Jigsaw-Solver

version - 1.0.0

This software solve the Jigsw puzzle using neural networks.

Requirements

  • python - 3.7
  • keras - 2.4.3
  • tensorflow - 2.2.0

Project Content

File Description
main.ipynb notebook for running and testing the code
model.py neural network implementation
utills.py contains helper functions
datagen_scripts scripts for generating data

Performance

Network was able to solve the puzzles with 80% accuracy. Following is a snippet showing network solving the puzzle.

References

  • Download the dataset for this project.
  • Read this blog to get all detail of this project.