Multi-shot video generation requires maintaining a consistent appearance of recurring entities across shots while remaining faithful to shot-specific text prompts. Recent autoregressive methods reuse previously generated frames as memory. However, full-frame storage entangles persistent entity information with transient scene context, leading to irrelevant information leakage and high computational cost. We propose an entity-centric memory in the form of an entity-indexed bank of latent patches. We introduce sparse token conditioning compatible with pretrained models, restricting self-attention to entity-relevant tokens and reducing computational cost. To support this, we introduce a structured multi-shot script format. We additionally propose a budgeted memory update strategy to maintain a compact, evolving memory. Finally, we equip the entity representation with a noise-injection mechanism that enables fine-grained appearance control, preventing leakage of irrelevant information. Our method improves prompt adherence and efficiency while preserving subject consistency.
@article{arxiv.2605.23610,
title = {EM-Vid: Training-Free Entity-Centric Memory for Efficient and Consistent Multi-Shot Video Generation},
author = {Jente Vandersanden and Matheus Gadelha and Chun-Hao P. Huang and Hyeonho Jeong and Yulia Gryaditskaya},
journal= {arXiv preprint arXiv:2605.23610},
year = {2026}
}