Scene Classification on Fine Arts with Style Transfer

Huang H, Zinnen M, Liu S, Maier A, Christlein V (2024)


Publication Type: Conference contribution

Publication year: 2024

Book Volume: 34

Pages Range: 18-27

Conference Proceedings Title: SUMAC '24: Proceedings of the 6th workshop on the analySis, Understanding and proMotion of heritAge Contents

DOI: 10.1145/3689094.3689468

Abstract

Large-scale photographic datasets like ImageNet and Places365 have significantly improved scene classification performance in natural images. However, scene classification in artistic imagery remains underexplored until recently. We propose a multi-step transfer learning technique that gradually adapts scene recognition algorithms from photographs to artistic scene representations. Our experiments demonstrate that integrating a stylized version of Places365, and fine-tuning with a weakly supervised artistic scene dataset, drastically increases scene recognition performance in artworks.We evaluate our method using two state-of-the-art scene recognition methods and analyze the impact of our adaptations with a series of ablation studies.

Authors with CRIS profile

How to cite

APA:

Huang, H., Zinnen, M., Liu, S., Maier, A., & Christlein, V. (2024). Scene Classification on Fine Arts with Style Transfer. In SUMAC '24: Proceedings of the 6th workshop on the analySis, Understanding and proMotion of heritAge Contents (pp. 18-27).

MLA:

Huang, Haiting, et al. "Scene Classification on Fine Arts with Style Transfer." Proceedings of the SUMAC '24: 6th workshop on the analySis, Understanding and proMotion of heritAge Contents 2024. 18-27.

BibTeX: Download