# QEVA: A Reference-Free Evaluation Metric for Narrative Video Summarization with Multimodal Question Answering

Source: [QEVA: A Reference-Free Evaluation Metric for Narrative Video Summarization with Multimodal Question Answering](<https://www.sota2.com/research/paper/69f03f2f55a3c32af89fa1df>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-29T05:00:12.215Z

Authors: Woojun Jung, Junyeong Kim

Published: 2026-04-27T05:18:21.000Z

## Abstract

Video-to-text summarization remains underexplored in terms of comprehensive evaluation methods. Traditional n-gram overlap-based metrics and recent large language model (LLM)-based approaches depend heavily on human-written reference summaries, limiting their practicality and sensitivity to nuanced semantic aspects. In this paper, we propose QEVA, a reference-free metric evaluating candidate summaries directly against source videos through multimodal question answering. QEVA assesses summaries along three clear dimensions: Coverage, Factuality, and Chronology. We also introduce MLVU(VS)-Eval, a new annotated benchmark derived from the MLVU dataset, comprising 800 summaries generated from 200 videos using state-of-the-art video-language multimodal models. This dataset establishes a transparent and consistent framework for evaluation. Experimental results demonstrate that QEVA shows higher correlation with human judgments compared to existing approaches, as measured by Kendall's $\\tau\_b$, $\\tau\_c$, and Spearman's $\\rho$. We hope that our benchmark and metric will facilitate meaningful progress in video-to-text summarization research and provide valuable insights for the development of future evaluation methods.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Video Summarization Metric Evaluation on MLVU VS Eval](<https://www.sota2.com/research/sota/video-summarization-metric-evaluation-on-mlvu-vs-eval.md>): Video Summarization Metric Evaluation; MLVU VS Eval; Kendall's tau\_b: 0.6465.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2604.24052>)

[PDF](<https://arxiv.org/pdf/2604.24052.pdf>)