Implementasi Vision Transformer (VIT) dan K-Means Clustering dalam Klasifikasi Gaya Visual Poster untuk Otomatisasi Ekstraksi Skema Warna Digital

Muhammadun Hanif; Bagus Satrio Waluyo Poetro; Ghufron Ghufron

doi:10.58192/populer.v5i3.5516

Authors

Muhammadun Hanif Universitas Islam Sultan Agung
Bagus Satrio Waluyo Poetro Universitas Islam Sultan Agung
Ghufron Ghufron Universitas Islam Sultan Agung

DOI:

https://doi.org/10.58192/populer.v5i3.5516

Keywords:

Color Scheme, Graphic Design, K-Means Clustering, Poster, Vision Transformer

Abstract

The purpose of this study is to implement the Vision Transformer (ViT) architecture and K-Means Clustering algorithm to automate fine-grained visual style classification of posters and the extraction of digital color schemes. MSMEs and beginner digital creators often face challenges in determining harmonious color schemes for promotional poster designs, leading to a time-consuming trial-and-error design process. The research utilized a custom dataset of 1,011 poster images scraped from the Behance platform, evenly distributed into three visual style classes: Bauhaus, Swiss Style, and Minimalist. The data preprocessing employed a dual-pipeline approach, where color extraction was optimized through SLIC Superpixel segmentation within the CIELab* perceptual color space to significantly reduce computational complexity without degrading the boundaries of original visual elements. The evaluation results demonstrated that the ViT model performed exceptionally well, achieving an accuracy of 79% and a Macro F1-Score of 0.80. In the color extraction module, the K-Means algorithm successfully formed five highly separated dominant palette clusters with an average Silhouette Score of 0.7698. The perceptual color accuracy validation recorded an average of 4.7648, proving that the extracted hexadecimal (HEX) palette codes operate well within the acceptable limits of human visual deviation.

References

Chen, Y. et al. (2021) “Investigation of designers’ colour selection process,” Color Research and Application, 46(3), hal. 557–565. Tersedia pada: https://doi.org/10.1002/col.22631.

Cherepkova, O., Amirshahi, S.A. dan Pedersen, M. (2024) “Individual Contrast Preferences in Natural Images,” Journal of Imaging, 10(1). Tersedia pada: https://doi.org/10.3390/jimaging10010025.

Conde, M. V. dan Turgutlu, K. (2021) “Exploring Vision Transformers for Fine-grained Classification,” (iii). Tersedia pada: http://arxiv.org/abs/2106.10587.

Dosovitskiy, A. et al. (2021) “an Image Is Worth 16X16 Words: Transformers for Image Recognition At Scale,” ICLR 2021 - 9th International Conference on Learning Representations [Preprint].

Frackiewicz, M. dan Palus, H. (2022) “Efficient Color Quantization Using Superpixels,” Sensors, 22(16). Tersedia pada: https://doi.org/10.3390/s22166043.

Frackiewicz, M., Palus, H. dan Prandzioch, D. (2023) “Superpixel-Based PSO Algorithms for Color Image Quantization,” Sensors, 23(3). Tersedia pada: https://doi.org/10.3390/s23031108.

Gao, Y., Liang, J. dan Yang, J. (2025) “Color Palette Generation From Digital Images: A Review,” Color Research and Application, 50(3), hal. 250–265. Tersedia pada: https://doi.org/10.1002/col.22975.

Garrison, L. dan Bruckner, S. (2022) “Considering best practices in color palettes for molecular visualizations,” Journal of integrative bioinformatics, 19(2), hal. 1–14. Tersedia pada: https://doi.org/10.1515/jib-2022-0016.

Han, A., Kim, J. dan Ahn, J. (2022) “Color Trend Analysis using Machine Learning with Fashion Collection Images,” Clothing and Textiles Research Journal, 40(4), hal. 308–324. Tersedia pada: https://doi.org/10.1177/0887302X21995948.

He, J. et al. (2022) “TransFG: A Transformer Architecture for Fine-Grained Recognition,” Proceedings of the 36th AAAI Conference on Artificial Intelligence, AAAI 2022, 36, hal. 1174–1182. Tersedia pada: https://doi.org/10.1609/aaai.v36i1.19967.

Hsu, H.Y. et al. (2023) “PosterLayout: A New Benchmark and Approach for Content-Aware Visual-Textual Presentation Layout,” Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2023-June, hal. 6018–6026. Tersedia pada: https://doi.org/10.1109/CVPR52729.2023.00583.

Huang, D. et al. (2023) “A Survey for Graphic Design Intelligence.” Tersedia pada: http://arxiv.org/abs/2309.01371.

Kösesoy, M.B. dan Yilmaz, S. (2025) “A Novel Color Difference-Based Method for Palette Extraction and Evaluation Using Images of Birds,” IEEE Access, 13(January), hal. 52270–52283. Tersedia pada: https://doi.org/10.1109/ACCESS.2025.3553252.

Liu, S. et al. (2024) “Image-Driven Harmonious Color Palette Generation for Diverse Information Visualization,” IEEE Transactions on Visualization and Computer Graphics, 30(7), hal. 3089–3103. Tersedia pada: https://doi.org/10.1109/TVCG.2022.3226218.

Sharma, S., Tandukar, J. dan Bista, R. (2023) “Generating Harmonious Colors through the Combination of n-Grams and K-means,” Journal of Computing Theories and Applications, 1(2), hal. 140–150. Tersedia pada: https://doi.org/10.33633/jcta.v1i2.9470.

Wang, M. et al. (2026) “An art style classification network integrating contrastive learning and counterfactual attention,” PLOS One, 21(6), hal. e0343954. Tersedia pada: https://doi.org/10.1371/journal.pone.0343954.

Wei, X.S. et al. (2022) “Fine-Grained Image Analysis With Deep Learning: A Survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(12), hal. 8927–8948. Tersedia pada: https://doi.org/10.1109/TPAMI.2021.3126648.

Yulisasih, B.N. et al. (2024) “Evaluation of K-Means Clustering Using Silhouette Score Method on Customer Segmentation,” ILKOM Jurnal Ilmiah, 16(3), hal. 330–342. Tersedia pada: https://doi.org/10.33096/ilkom.v16i3.2325.330-342.