MuseBench: Benchmarking Intent-Level Audiovisual Arts Understanding in MLLMs

• URL: http://arxiv.org/abs/2606.30026v1
• Date: Mon, 29 Jun 2026 09:27:02 GMT
• Title: MuseBench: Benchmarking Intent-Level Audiovisual Arts Understanding in MLLMs
• Abstract summary: Musebench is a benchmark designed to evaluate large language models (MLLMs) on nuanced artistic understanding.<n>It comprises 4,016 questions spanning cinematic arts, static visual arts, stage performing arts, and game arts.<n>A zero-shot evaluation of 28 state-of-the-art MLLMs reveals that even the best-performing model achieves only 48.29% accuracy.
• Score: 66.28927610631096
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Audiovisual arts encompass diverse creative disciplines, including cinema, visual arts, stage performance, and game design, where artistic meaning arises from deliberate combinations of visual, auditory, and narrative elements (e.g., fear amplified through claustrophobic framing, or grief conveyed through silence and lingering close-ups). True artistic understanding extends beyond recognizing what is depicted to reasoning about why it is expressed through particular creative choices. Despite the strong progress of multimodal large language models (MLLMs), this critical aspect of artistic understanding remains underexplored, as existing benchmarks largely measure perceptual recognition while overlooking reasoning about creative intent. To address this gap, we introduce Musebench, a comprehensive benchmark designed to evaluate MLLMs on nuanced artistic understanding. It comprises 4,016 questions spanning cinematic arts, static visual arts, stage performing arts, and game arts, distilled from over 10K candidate video essays that pair professional commentary with visual demonstration. To capture the open-ended nature of artistic analysis at scale, the benchmark combines single-select and variable-option multi-select questions. All questions are generated and refined through a four-phase iterative pipeline combining shortcut filtering, adversarial distractors, and expert validation. Comprehensive zero-shot evaluation of 28 state-of-the-art MLLMs reveals that even the best-performing model achieves only 48.29% accuracy, substantially below human expert performance of 87.18%, exposing a significant gap in current models' creative domain expertise.

Related papers

• Does AI See like Art Historians? Interpreting How Vision Language Models Recognize Artistic Style [60.562013130566726]
  We characterize the mechanisms underlying VLMs' ability to predict artistic style.<n>We employ a latent-space decomposition approach to identify concepts that drive art style prediction.<n>In cases where an irrelevant concept was used to successfully predict style, art historians identified possible reasons for its success.
  arXiv  Detail & Related papers  (2026-03-11T17:49:45Z)
• Does CLIP perceive art the same way we do? [0.05833117322405446]
  We investigate CLIP's ability to extract high-level semantic and stylistic information from paintings.<n>Our findings reveal both strengths and limitations in CLIP's visual representations.<n>Our work highlights the need for deeper interpretability in multimodal systems.
  arXiv  Detail & Related papers  (2025-05-08T13:21:10Z)
• CognArtive: Large Language Models for Automating Art Analysis and Decoding Aesthetic Elements [1.0579965347526206]
  Art, as a universal language, can be interpreted in diverse ways.<n>Large Language Models (LLMs) and the availability of Multimodal Large Language Models (MLLMs) raise the question of how these models can be used to assess and interpret artworks.
  arXiv  Detail & Related papers  (2025-02-04T18:08:23Z)
• Multimodal LLMs Can Reason about Aesthetics in Zero-Shot [35.59051707152096]
  This paper investigates how the reasoning capabilities of Multimodal LLMs can be effectively elicited to perform aesthetic judgment.<n>We show that MLLMs exhibit a tendency towards hallucinations during aesthetic reasoning, characterized by subjective opinions and unsubstantiated artistic interpretations.<n> MLLMs prompted by this principle produce multifaceted, in-depth aesthetic reasoning that aligns significantly better with human judgment.
  arXiv  Detail & Related papers  (2025-01-15T18:56:22Z)
• AesBench: An Expert Benchmark for Multimodal Large Language Models on Image Aesthetics Perception [64.25808552299905]
  AesBench is an expert benchmark aiming to comprehensively evaluate the aesthetic perception capacities of MLLMs. We construct an Expert-labeled Aesthetics Perception Database (EAPD), which features diversified image contents and high-quality annotations provided by professional aesthetic experts. We propose a set of integrative criteria to measure the aesthetic perception abilities of MLLMs from four perspectives, including Perception (AesP), Empathy (AesE), Assessment (AesA) and Interpretation (AesI)
  arXiv  Detail & Related papers  (2024-01-16T10:58:07Z)
• Impressions: Understanding Visual Semiotics and Aesthetic Impact [66.40617566253404]
  We present Impressions, a novel dataset through which to investigate the semiotics of images. We show that existing multimodal image captioning and conditional generation models struggle to simulate plausible human responses to images. This dataset significantly improves their ability to model impressions and aesthetic evaluations of images through fine-tuning and few-shot adaptation.
  arXiv  Detail & Related papers  (2023-10-27T04:30:18Z)
• A Video Is Worth 4096 Tokens: Verbalize Videos To Understand Them In Zero Shot [67.00455874279383]
  We propose verbalizing long videos to generate descriptions in natural language, then performing video-understanding tasks on the generated story as opposed to the original video. Our method, despite being zero-shot, achieves significantly better results than supervised baselines for video understanding. To alleviate a lack of story understanding benchmarks, we publicly release the first dataset on a crucial task in computational social science on persuasion strategy identification.
  arXiv  Detail & Related papers  (2023-05-16T19:13:11Z)
• Art Style Classification with Self-Trained Ensemble of AutoEncoding Transformations [5.835728107167379]
  Artistic style of a painting is a rich descriptor that reveals both visual and deep intrinsic knowledge about how an artist uniquely portrays and expresses their creative vision. In this paper, we investigate the use of deep self-supervised learning methods to solve the problem of recognizing complex artistic styles with high intra-class and low inter-class variation.
  arXiv  Detail & Related papers  (2020-12-06T21:05:23Z)
• Multi-Modal Music Information Retrieval: Augmenting Audio-Analysis with Visual Computing for Improved Music Video Analysis [91.3755431537592]
  This thesis combines audio-analysis with computer vision to approach Music Information Retrieval (MIR) tasks from a multi-modal perspective. The main hypothesis of this work is based on the observation that certain expressive categories such as genre or theme can be recognized on the basis of the visual content alone. The experiments are conducted for three MIR tasks Artist Identification, Music Genre Classification and Cross-Genre Classification.
  arXiv  Detail & Related papers  (2020-02-01T17:57:14Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.