ImagerySearch: Adaptive Test-Time Search for Video Generation Beyond Semantic Dependency Constraints

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Wu, Meiqi, Zhu, Jiashu, Feng, Xiaokun, Chen, Chubin, Zhu, Chen, Song, Bingze, Mao, Fangyuan, Wu, Jiahong, Chu, Xiangxiang, Huang, Kaiqi
Format:	Preprint
Published:	2025
Subjects:	Computer Vision and Pattern Recognition
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

_version_	1866915569941348352
author	Wu, Meiqi Zhu, Jiashu Feng, Xiaokun Chen, Chubin Zhu, Chen Song, Bingze Mao, Fangyuan Wu, Jiahong Chu, Xiangxiang Huang, Kaiqi
author_facet	Wu, Meiqi Zhu, Jiashu Feng, Xiaokun Chen, Chubin Zhu, Chen Song, Bingze Mao, Fangyuan Wu, Jiahong Chu, Xiangxiang Huang, Kaiqi
contents	Video generation models have achieved remarkable progress, particularly excelling in realistic scenarios; however, their performance degrades notably in imaginative scenarios. These prompts often involve rarely co-occurring concepts with long-distance semantic relationships, falling outside training distributions. Existing methods typically apply test-time scaling for improving video quality, but their fixed search spaces and static reward designs limit adaptability to imaginative scenarios. To fill this gap, we propose ImagerySearch, a prompt-guided adaptive test-time search strategy that dynamically adjusts both the inference search space and reward function according to semantic relationships in the prompt. This enables more coherent and visually plausible videos in challenging imaginative settings. To evaluate progress in this direction, we introduce LDT-Bench, the first dedicated benchmark for long-distance semantic prompts, consisting of 2,839 diverse concept pairs and an automated protocol for assessing creative generation capabilities. Extensive experiments show that ImagerySearch consistently outperforms strong video generation baselines and existing test-time scaling approaches on LDT-Bench, and achieves competitive improvements on VBench, demonstrating its effectiveness across diverse prompt types. We will release LDT-Bench and code to facilitate future research on imaginative video generation.
format	Preprint
id	arxiv_https___arxiv_org_abs_2510_14847
institution	arXiv
publishDate	2025
record_format	arxiv
spellingShingle	ImagerySearch: Adaptive Test-Time Search for Video Generation Beyond Semantic Dependency Constraints Wu, Meiqi Zhu, Jiashu Feng, Xiaokun Chen, Chubin Zhu, Chen Song, Bingze Mao, Fangyuan Wu, Jiahong Chu, Xiangxiang Huang, Kaiqi Computer Vision and Pattern Recognition Video generation models have achieved remarkable progress, particularly excelling in realistic scenarios; however, their performance degrades notably in imaginative scenarios. These prompts often involve rarely co-occurring concepts with long-distance semantic relationships, falling outside training distributions. Existing methods typically apply test-time scaling for improving video quality, but their fixed search spaces and static reward designs limit adaptability to imaginative scenarios. To fill this gap, we propose ImagerySearch, a prompt-guided adaptive test-time search strategy that dynamically adjusts both the inference search space and reward function according to semantic relationships in the prompt. This enables more coherent and visually plausible videos in challenging imaginative settings. To evaluate progress in this direction, we introduce LDT-Bench, the first dedicated benchmark for long-distance semantic prompts, consisting of 2,839 diverse concept pairs and an automated protocol for assessing creative generation capabilities. Extensive experiments show that ImagerySearch consistently outperforms strong video generation baselines and existing test-time scaling approaches on LDT-Bench, and achieves competitive improvements on VBench, demonstrating its effectiveness across diverse prompt types. We will release LDT-Bench and code to facilitate future research on imaginative video generation.
title	ImagerySearch: Adaptive Test-Time Search for Video Generation Beyond Semantic Dependency Constraints
topic	Computer Vision and Pattern Recognition
url	https://arxiv.org/abs/2510.14847