StoryMem

--- base_model: - Wan-AI/Wan2.2-I2V-A14B license: cc-by-nc-4.0 pipeline_tag: text-to-video --- <p align="center"> <h2 align="center"><i>StoryMem</i>: Multi-shot Long Video Storytelling with Memory</h1> <p align="center"> <span class="author-block"> <a href="https://kevin-thu.github.io/homepage/" target="_blank">Kaiwen Zhang</a>, </span> <span class="author-block"> <a href="https://liming-jiang.com/" target="_blank">Liming Jiang</a>^✝, </span> <span class="author-block"> <a href="https://angtianwang.github.io/" target="_blank">Angtian Wang</a>, </span> <span class="author-block"> <a href="https://www.jacobzhiyuanfang.me/" target="_blank">Jacob Zhiyuan Fang</a>, </span><br> <span class="author-block"> <a href="https://tiancheng-zhi.github.io/" target="_blank">Tiancheng Zhi</a>, </span> <span class="author-block"> <a href="https://scholar.google.com/citations?user=0TIYjPAAAAAJ&hl=en" target="_blank">Qing Yan</a>, </span> <span class="author-block"> <a href="https://scholar.google.com/citations?user=VeTCSyEAAAAJ" target="_blank">Hao Kang</a>, </span> <span class="author-block"> <a href="https://scholar.google.com/citations?user=mFC0wp8AAAAJ&hl=en" target="_blank">Xin Lu</a>, </span> <span class="author-block"> <a href="https://xingangpan.github.io/" target="_blank">Xingang Pan</a>^§ </p> <p align="center"> <sep>✝</sep>Project Lead &nbsp; <sep>§</sep>Corresponding Author </p> <p align="center"> <a href="https://arxiv.org/abs/2512.19539"><img alt='arXiv' src="https://img.shields.io/badge/arXiv-2512.19539-b31b1b.svg"></a> <a href="https://kevin-thu.github.io/StoryMem/"><img alt='page' src="https://img.shields.io/badge/Project-Website-orange"></a> <a href="https://huggingface.co/Kevin-thu/StoryMem"><img alt="Huggingface" src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Model-orange"></a> <!-- <a href="https://twitter.com/sze68zkw"><img alt='Twitter' src="https://img.shields.io/twitter/follow/sze68zkw?label=%40KaiwenZhang"></a> --> </p> <div align="center"> <img src="./assets/teaser.png", width="800"> <p align="left">Given a story script with per-shot text descriptions, StoryMem generates appealing minute-long, multi-shot narrative videos with highly coherent characters and cinematic visual quality. This is achieved through shot-by-shot generation using a memory-conditioned single-sho...

Source context: 22 downloads · 97 likes · Pipeline text-to-video · Repo Kevin-thu/StoryMem