MA-Bench: Towards Fine-grained Micro-Action Understanding

TL;DR

MA-Bench introduces a comprehensive benchmark and training dataset for evaluating and improving multimodal large language models in fine-grained micro-action understanding and human behavior analysis.

Contribution

The paper presents MA-Bench, a new benchmark with structured questions and a training corpus for advancing micro-action perception in multimodal models.

Findings

Existing MLLMs struggle with motion granularity and body-part dynamics.

Fine-tuning on MA-Bench-Train improves micro-action reasoning.

Benchmark enables systematic assessment of recognition and interpretation.

Abstract

With the rapid development of Multimodal Large Language Models (MLLMs), their potential in Micro-Action understanding, a vital role in human emotion analysis, remains unexplored due to the absence of specialized benchmarks. To tackle this issue, we present MA-Bench, a benchmark comprising 1,000 videos and a three-tier evaluation architecture that progressively examines micro-action perception, relational comprehension, and interpretive reasoning. MA-Bench contains 12,000 structured question-answer pairs, enabling systematic assessment of both recognition accuracy and action interpretation. The results of 23 representative MLLMs reveal that there are significant challenges in capturing motion granularity and fine-grained body-part dynamics. To address these challenges, we further construct MA-Bench-Train, a large-scale training corpus with 20.5K videos annotated with structured micro-action captions for fine-tuning MLLMs. The results of Qwen3-VL-8B fine-tuned on MA-Bench-Train show clear performance improvements across micro-action reasoning and explanation tasks. Our work aims to establish a foundation benchmark for advancing MLLMs in understanding subtle micro-action and human-related behaviors. Project Page: https://MA-Bench.github.io