Saved in:
Bibliographic Details
Main Authors: Wu, Xinyi, Teotia, Jayant, Zhao, Shuai, Cambria, Erik
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.09378
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909030133268480
author Wu, Xinyi
Teotia, Jayant
Zhao, Shuai
Cambria, Erik
author_facet Wu, Xinyi
Teotia, Jayant
Zhao, Shuai
Cambria, Erik
contents Long-horizon video generation has advanced in visual quality, yet existing methods still struggle to maintain knowledge consistency and coherent pedagogical narratives across multi-shot instructional videos, especially in STEM domains. To address these challenges, we propose EduStory, a unified framework for reliable instructional video generation. EduStory integrates pedagogical state modeling to track persistent knowledge states, script-guided structured control to organize multi-shot narratives, and learning-oriented evaluation metrics to assess knowledge fidelity and constraint satisfaction. To support rigorous evaluation, we further introduce EduVideoBench, a diagnostic benchmark with multi-granularity annotations, including pedagogical storyboards, shot-level semantics, and knowledge state transitions, together with baseline tasks for controllable instructional video generation. Extensive experiments demonstrate that domain-aware state modeling and structured control substantially reduce narrative breakdown and improve alignment with instructional intent. These results highlight the significance of domain-specific structural constraints and tailored benchmarks for advancing reliable, controllable, and also trustworthy long-horizon video generation.
format Preprint
id arxiv_https___arxiv_org_abs_2605_09378
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle EduStory: A Unified Framework for Pedagogically-Consistent Multi-Shot STEM Instructional Video Generation
Wu, Xinyi
Teotia, Jayant
Zhao, Shuai
Cambria, Erik
Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
Long-horizon video generation has advanced in visual quality, yet existing methods still struggle to maintain knowledge consistency and coherent pedagogical narratives across multi-shot instructional videos, especially in STEM domains. To address these challenges, we propose EduStory, a unified framework for reliable instructional video generation. EduStory integrates pedagogical state modeling to track persistent knowledge states, script-guided structured control to organize multi-shot narratives, and learning-oriented evaluation metrics to assess knowledge fidelity and constraint satisfaction. To support rigorous evaluation, we further introduce EduVideoBench, a diagnostic benchmark with multi-granularity annotations, including pedagogical storyboards, shot-level semantics, and knowledge state transitions, together with baseline tasks for controllable instructional video generation. Extensive experiments demonstrate that domain-aware state modeling and structured control substantially reduce narrative breakdown and improve alignment with instructional intent. These results highlight the significance of domain-specific structural constraints and tailored benchmarks for advancing reliable, controllable, and also trustworthy long-horizon video generation.
title EduStory: A Unified Framework for Pedagogically-Consistent Multi-Shot STEM Instructional Video Generation
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2605.09378