ViMU: Benchmark for Video Metaphorical Understanding

ViMU is a benchmark for evaluating multimodal models on video metaphorical understanding. The code repository from the National University of Singapore contains evaluation scripts for four distinct tasks. The dataset page was last updated on 2026-05-16.

Use Cases

• Benchmarking model performance on open-ended video interpretation tasks based on the described evaluation framework.
• Evaluating evidence grounding capabilities in multimodal models based on the described task.
• Identifying rhetorical mechanisms in video content based on the described benchmark task.
• Analyzing social value signals in video narratives based on the described evaluation category.

Strengths

• Defines four specific evaluation tasks for video metaphorical understanding: Open-ended interpretation, Evidence grounding, Rhetoric mechanism identification, and Social value signal identification.
• Associated with a research institution, the National University of Singapore.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and file formats are unknown, which may limit suitability assessment.

Provenance

Source

LIQIIIII on Hugging Face, associated with the xML Lab, National University of Singapore.

Collection Method

Likely created as a research benchmark for evaluating AI models.

Freshness

Last updated 2026-05-16 12:23:52; freshness should be verified.