VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ma, Wentao, Ren, Weiming, Jia, Yiming, Li, Zhuofeng, Nie, Ping, Zhang, Ge, Chen, Wenhu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918027029643264
author Ma, Wentao
Ren, Weiming
Jia, Yiming
Li, Zhuofeng
Nie, Ping
Zhang, Ge
Chen, Wenhu
author_facet Ma, Wentao
Ren, Weiming
Jia, Yiming
Li, Zhuofeng
Nie, Ping
Zhang, Ge
Chen, Wenhu
contents Large multimodal models (LMMs) have recently emerged as a powerful tool for long video understanding (LVU), prompting the development of standardized LVU benchmarks to evaluate their performance. However, our investigation reveals a rather sober lesson for existing LVU benchmarks. First, most existing benchmarks rely heavily on multiple-choice questions (MCQs), whose evaluation results are inflated due to the possibility of guessing the correct answer; Second, a significant portion of questions in these benchmarks have strong priors to allow models to answer directly without even reading the input video. For example, Gemini-1.5-Pro can achieve over 50\% accuracy given a random frame from a long video on Video-MME. We also observe that increasing the number of frames does not necessarily lead to improvement on existing benchmarks, which is counterintuitive. As a result, the validity and robustness of current LVU benchmarks are undermined, impeding a faithful assessment of LMMs' long-video understanding capability. To tackle this problem, we propose VideoEval-Pro, a realistic LVU benchmark containing questions with open-ended short-answer, which truly require understanding the entire video. VideoEval-Pro assesses both segment-level and full-video understanding through perception and reasoning tasks. By evaluating 21 proprietary and open-source video LMMs, we conclude the following findings: (1) video LMMs show drastic performance ($>$25\%) drops on open-ended questions compared with MCQs; (2) surprisingly, higher MCQ scores do not lead to higher open-ended scores on VideoEval-Pro; (3) compared to other MCQ benchmarks, VideoEval-Pro benefits more from increasing the number of input frames. Our results show that VideoEval-Pro offers a more realistic and reliable measure of long video understanding, providing a clearer view of progress in this domain.
format Preprint
id arxiv_https___arxiv_org_abs_2505_14640
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation
Ma, Wentao
Ren, Weiming
Jia, Yiming
Li, Zhuofeng
Nie, Ping
Zhang, Ge
Chen, Wenhu
Computer Vision and Pattern Recognition
Large multimodal models (LMMs) have recently emerged as a powerful tool for long video understanding (LVU), prompting the development of standardized LVU benchmarks to evaluate their performance. However, our investigation reveals a rather sober lesson for existing LVU benchmarks. First, most existing benchmarks rely heavily on multiple-choice questions (MCQs), whose evaluation results are inflated due to the possibility of guessing the correct answer; Second, a significant portion of questions in these benchmarks have strong priors to allow models to answer directly without even reading the input video. For example, Gemini-1.5-Pro can achieve over 50\% accuracy given a random frame from a long video on Video-MME. We also observe that increasing the number of frames does not necessarily lead to improvement on existing benchmarks, which is counterintuitive. As a result, the validity and robustness of current LVU benchmarks are undermined, impeding a faithful assessment of LMMs' long-video understanding capability. To tackle this problem, we propose VideoEval-Pro, a realistic LVU benchmark containing questions with open-ended short-answer, which truly require understanding the entire video. VideoEval-Pro assesses both segment-level and full-video understanding through perception and reasoning tasks. By evaluating 21 proprietary and open-source video LMMs, we conclude the following findings: (1) video LMMs show drastic performance ($>$25\%) drops on open-ended questions compared with MCQs; (2) surprisingly, higher MCQ scores do not lead to higher open-ended scores on VideoEval-Pro; (3) compared to other MCQ benchmarks, VideoEval-Pro benefits more from increasing the number of input frames. Our results show that VideoEval-Pro offers a more realistic and reliable measure of long video understanding, providing a clearer view of progress in this domain.
title VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2505.14640