Vision-Language Models Show Temporal Knowledge in Art Dating, But Biases Remain

Researchers have developed a method for dating artworks using vision-language models (VLMs), addressing the issue of "temporal entanglement" where models appear to encode historical time but actually reflect institutional biases in data collection. The study formulated artwork dating as an uncertainty-aware regression task using frozen image embeddings. Results on a Wikidata corpus showed that VLMs could extract usable temporal information, outperforming purely visual models, though a qualitative analysis revealed biases in their temporal knowledge. AI

IMPACT This research highlights the potential of VLMs for historical analysis while cautioning about inherent biases in their temporal understanding.

RANK_REASON The cluster contains an academic paper detailing a new methodology and findings in computer vision. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• arXiv
• Hugging Face
• Stefanie Schneider
• vision-language model
• Wikidata

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →