Saved in:
Bibliographic Details
Main Authors: Wu, Jay Zhangjie, Fang, Guian, Wu, Haoning, Wang, Xintao, Ge, Yixiao, Cun, Xiaodong, Zhang, David Junhao, Liu, Jia-Wei, Gu, Yuchao, Zhao, Rui, Lin, Weisi, Hsu, Wynne, Shan, Ying, Shou, Mike Zheng
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2401.07781
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909073571577856
author Wu, Jay Zhangjie
Fang, Guian
Wu, Haoning
Wang, Xintao
Ge, Yixiao
Cun, Xiaodong
Zhang, David Junhao
Liu, Jia-Wei
Gu, Yuchao
Zhao, Rui
Lin, Weisi
Hsu, Wynne
Shan, Ying
Shou, Mike Zheng
author_facet Wu, Jay Zhangjie
Fang, Guian
Wu, Haoning
Wang, Xintao
Ge, Yixiao
Cun, Xiaodong
Zhang, David Junhao
Liu, Jia-Wei
Gu, Yuchao
Zhao, Rui
Lin, Weisi
Hsu, Wynne
Shan, Ying
Shou, Mike Zheng
contents Generative models have demonstrated remarkable capability in synthesizing high-quality text, images, and videos. For video generation, contemporary text-to-video models exhibit impressive capabilities, crafting visually stunning videos. Nonetheless, evaluating such videos poses significant challenges. Current research predominantly employs automated metrics such as FVD, IS, and CLIP Score. However, these metrics provide an incomplete analysis, particularly in the temporal assessment of video content, thus rendering them unreliable indicators of true video quality. Furthermore, while user studies have the potential to reflect human perception accurately, they are hampered by their time-intensive and laborious nature, with outcomes that are often tainted by subjective bias. In this paper, we investigate the limitations inherent in existing metrics and introduce a novel evaluation pipeline, the Text-to-Video Score (T2VScore). This metric integrates two pivotal criteria: (1) Text-Video Alignment, which scrutinizes the fidelity of the video in representing the given text description, and (2) Video Quality, which evaluates the video's overall production caliber with a mixture of experts. Moreover, to evaluate the proposed metrics and facilitate future improvements on them, we present the TVGE dataset, collecting human judgements of 2,543 text-to-video generated videos on the two criteria. Experiments on the TVGE dataset demonstrate the superiority of the proposed T2VScore on offering a better metric for text-to-video generation.
format Preprint
id arxiv_https___arxiv_org_abs_2401_07781
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Towards A Better Metric for Text-to-Video Generation
Wu, Jay Zhangjie
Fang, Guian
Wu, Haoning
Wang, Xintao
Ge, Yixiao
Cun, Xiaodong
Zhang, David Junhao
Liu, Jia-Wei
Gu, Yuchao
Zhao, Rui
Lin, Weisi
Hsu, Wynne
Shan, Ying
Shou, Mike Zheng
Computer Vision and Pattern Recognition
Generative models have demonstrated remarkable capability in synthesizing high-quality text, images, and videos. For video generation, contemporary text-to-video models exhibit impressive capabilities, crafting visually stunning videos. Nonetheless, evaluating such videos poses significant challenges. Current research predominantly employs automated metrics such as FVD, IS, and CLIP Score. However, these metrics provide an incomplete analysis, particularly in the temporal assessment of video content, thus rendering them unreliable indicators of true video quality. Furthermore, while user studies have the potential to reflect human perception accurately, they are hampered by their time-intensive and laborious nature, with outcomes that are often tainted by subjective bias. In this paper, we investigate the limitations inherent in existing metrics and introduce a novel evaluation pipeline, the Text-to-Video Score (T2VScore). This metric integrates two pivotal criteria: (1) Text-Video Alignment, which scrutinizes the fidelity of the video in representing the given text description, and (2) Video Quality, which evaluates the video's overall production caliber with a mixture of experts. Moreover, to evaluate the proposed metrics and facilitate future improvements on them, we present the TVGE dataset, collecting human judgements of 2,543 text-to-video generated videos on the two criteria. Experiments on the TVGE dataset demonstrate the superiority of the proposed T2VScore on offering a better metric for text-to-video generation.
title Towards A Better Metric for Text-to-Video Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2401.07781