Skip to content

Source paper

StoryMem: Multi-shot Long Video Storytelling with Memory

Kaiwen Zhang, Liming Jiang, Angtian Wang and 6 more

Kaiwen Zhang, Liming Jiang, Angtian Wang, Jacob Zhiyuan Fang, Tiancheng Zhi, Qing Yan, Hao Kang, Xin Lu, Xingang Pan

arXiv 2512.19539 · 202514 upvotes on Hugging Face

Research ideas from this paper (3)

Generated by AI. The paper’s authors did not write these.

AI summary

The paper introduces StoryMem, a framework that enables pre-trained single-shot video diffusion models to generate coherent, multi-shot long videos. It employs a Memory-to-Video (M2V) mechanism that stores and retrieves visual keyframes from previous shots to maintain narrative consistency across minute-long sequences.

How it works (AI summary)

The method adapts the Wan2.2-I2V diffusion transformer (DiT) using LoRA fine-tuning. It utilizes a dynamic memory bank of keyframes, selected via CLIP semantic similarity and HPSv3 aesthetic filtering, which are injected into the generation process through latent concatenation and a novel negative Rotary Positional Embedding (RoPE) shift strategy to represent past context.