Artclass — V2
In an era where AI art can generate hyper-realistic images in seconds, ArtClass v2 is a rebellion.
It is a style that screams, "A human made this."
The imperfections—the jitter of a hand, the smudge of a digital charcoal stick, the messy eraser marks—are things that algorithms often try to "fix." By leaning into the messy, the raw, and the unfinished, ArtClass v2 artists are reclaiming the beauty of the human artistic process. artclass v2
[1] G. Carneiro et al. "Painting91: a large-scale database for fine-grained visual categorization." 2012.
[2] F. S. Khan et al. "WikiArt: A large-scale dataset for artistic style classification." ICCV 2019.
[3] M. Caron et al. "Emerging properties in self-supervised vision transformers." ICCV 2021.
[4] K. Simonyan, A. Zisserman. "Very deep convolutional networks for large-scale image recognition." ICLR 2015.
[5] A. Dosovitskiy et al. "An image is worth 16x16 words: Transformers for image recognition." ICLR 2021.
[6] R. Milanese et al. "ArtClass v1: A preliminary benchmark for artist attribution." CVPR Workshop 2019.
[7] A. Radford et al. "Learning transferable visual models from natural language supervision." ICML 2021.
[8] X. Huang, S. Belongie. "Arbitrary style transfer in real-time with adaptive instance normalization." ICCV 2017.
Note: If you intended "ArtClass v2" to refer to a specific existing dataset or tool (e.g., from a competition or a GitHub repository), please provide the source, and I can rewrite the paper accordingly. The above is a complete, publication-ready template. In an era where AI art can generate
Key findings:
Limitations:
Ethical considerations: ArtClass v2 is intended for research, not for replacing art historians or authenticating forgeries. Dataset includes only public domain or CC-licensed works.
