Research map / Vision-language and multimodal / Human and embodied modeling
ArtEmis: Affective Language for Visual Art
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add the predicted emotion label as a generation condition (ablation) | SAT speaker · ArtEmis test | Emo-Align: 0.400 -> 0.522 | Helped |
| Add the predicted emotion label as a generation condition (ablation) | SAT speaker · ArtEmis test | BLEU-1: 0.505 -> 0.505 | No clear effect |
| Add the predicted emotion label as a generation condition (ablation) | SAT speaker · ArtEmis test | BLEU-4: 0.081 -> 0.080 | Hurt |
| Train the speaker on ArtEmis rather than relying on nearest-neighbor captions or sentiment adjective injection | ArtEmis neural speakers · ArtEmis test | BLEU-1: 0.346 -> 0.484; 0.386 -> 0.484 | Helped |
| Use emotion grounding in the SAT speaker (ablation) | ArtEmis neural speaker · ArtEmis emotional Turing test | human-parity-or-better judgments: 40% -> 50.3% | Helped |
| Use emotion grounding in the SAT speaker (ablation) | ArtEmis neural speaker · ArtEmis test | Similes-percent: 0.452 -> 0.356; ground truth: ~20.5 | Mixed |
Closest papers by meaning
- Customizing Visual Emotion Evaluation for MLLMs: An Open-vocabulary, Multifaceted, and Scalable Approach (ICLR 2026 · similarity 0.87)
- EmoEdit: Evoking Emotions through Image Manipulation (CVPR 2025 · similarity 0.87)
- EmoGen: Emotional Image Content Generation with Text-to-Image Diffusion Models (CVPR 2024 · similarity 0.87)
- Bridge Then Begin Anew: Generating Target-Relevant Intermediate Model for Source-Free Visual Emotion Adaptation (AAAI 2025 · similarity 0.86)
- Bridging Cognitive Gap: Hierarchical Description Learning for Artistic Image Aesthetics Assessment (AAAI 2026 · similarity 0.85)
- Gender Artifacts from Art History to Text-to-Image Generation (NeurIPS 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.