# ArtSociety: Multi-Agent Multimodal Collaboration for Art Emotion Understanding

arXiv: https://arxiv.org/abs/2609.13240 · HTML: https://mlsanity.com/paper/2609.13240/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Multimodal and embodied agents, one of 16 directions in LLM agents and tool use.
- Builds on 12 of its 31 references that are in the map.
- First posted on arXiv 3 Sep 2026.

## Placed in

- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 12 of 12)

- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Improving Factuality and Reasoning in Language Models through Multiagent Debate](https://mlsanity.com/paper/2305.14325/) (2023)
- [Class-Balanced Loss Based on Effective Number of Samples](https://mlsanity.com/paper/1901.05555/) (2019)
- [AffectNet: A Database for Facial Expression, Valence, and Arousal Computing in the Wild](https://mlsanity.com/paper/1708.03985/) (2017)
- [Mixture-of-Agents Enhances Large Language Model Capabilities](https://mlsanity.com/paper/2406.04692/) (2024)
- [ArtEmis: Affective Language for Visual Art](https://mlsanity.com/paper/2101.07396/) (2021)
- [EmoSet: A Large-scale Visual Emotion Dataset with Rich Attributes](https://mlsanity.com/paper/2307.07961/) (2023)
- [Building a Large Scale Dataset for Image Emotion Recognition: The Fine Print and The Benchmark](https://mlsanity.com/paper/1605.02677/) (2016)
- [It is Okay to Not Be Okay: Overcoming Emotional Bias in Affective Image Captioning by Contrastive Data Collection](https://mlsanity.com/paper/2204.07660/) (2022)
- [FindingEmo: An Image Dataset for Emotion Recognition in the Wild](https://mlsanity.com/paper/2402.01355/) (2024)
- [EmoArt: A Multidimensional Dataset for Emotion-Aware Artistic Generation](https://mlsanity.com/paper/2506.03652/) (2025)
- [EmoVLM-KD: Fusing Distilled Expertise with Vision-Language Models for Visual Emotion Analysis](https://mlsanity.com/paper/2505.07164/) (2025)

## Abstract (arXiv)

The AffectiveArt Multidimensional Art Emotion Understanding task asks to jointly predict an artwork's fine-grained emotion (12 classes, 1549:1 head-to-tail ratio), binary valence/arousal, and five attribute-grounded descriptions -- sub-tasks that exhibit strong empirical trade-offs, so the single-model solutions we tried do not jointly optimize all of them well. We present ArtSociety, a multi-agent framework that assembles heterogeneous multimodal experts -- a DINOv2-Giant vision agent (A1), a scene-grounded CoT fine-tuned MLLM (A2), and three closed-source reasoning agents (A3-A5) -- and coordinates them with two training-free controllers: (i) a rare-class-aware voting arbiter that lowers the agreement threshold for tail emotions, exploiting decorrelated error patterns across agent families; and (ii) a description-first reasoning agent whose DESCRIBE-then-CLASSIFY chain of thought forces visual evidence before label commitment, yielding near-perfect grounded descriptions. A task-routing policy directs the hard emotion task to the full five-agent ensemble while assigning the near-saturated valence/arousal and generative description tasks to the single strongest reasoning agent. On the official test set (1,000 artworks), ArtSociety achieves an Overall Score of 0.8870 (Classification 0.7789, Description 0.9952). An eleven-variant ablation study reveals that, once method and scale saturate at around 0.76, the decisive gains come from agent collaboration and data-side supervision -- a 30B MoE model trained on older data does not outperform an 8B model trained on better data. Code is available at https://github.com/swordlidev/ArtSociety