Batik-Classification

(2 Oct 2021) Happy Batik Day!

A short introduction to Indonesian batik classification using CNN by NgodingPython during his livestream session.

According to Yohanes Gultom, there are 5 classes in the dataset, namely:

1. Parang: parang (traditional blade) motifs
2. Lereng: also blade-like pattern but less pointy than Parang
3. Ceplok: repetition of geometrical motif shapes (e.g. rectangle, square, ovals, stars)
4. Kawung: kawung fruits motif
5. Nitik: flowers-like motifs

This means the sixth class, Mix motif, is just a mixture of these five.

References

Dataset can be found here. Full resources from Yohanes Gultom.

@article{gultom2018batik,
  title={Batik Classification using Deep Convolutional Network Transfer Learning},
  author={Gultom, Yohanes and Arymurthy, Aniati Murni and Masikome, Rian Josua},
  journal={Jurnal Ilmu Komputer dan Informasi},
  volume={11},
  number={2},
  pages={59--66},
  year={2018}
}

Note to Self

Additional resources

• https://www.pyimagesearch.com/blog/
• https://www.deeplearningbook.org/