VideoMemory: Toward Consistent Video Generation via Memory Integration
This addresses the problem of inconsistent entity identity in long-form video generation for applications like storytelling and filmmaking, representing a novel method for a known bottleneck.
The paper tackles the challenge of maintaining consistent characters, props, and environments in narrative video generation by introducing VideoMemory, a framework that integrates a Dynamic Memory Bank to store and update entity representations, achieving strong entity-level coherence and high perceptual quality across diverse sequences.
Maintaining consistent characters, props, and environments across multiple shots is a central challenge in narrative video generation. Existing models can produce high-quality short clips but often fail to preserve entity identity and appearance when scenes change or when entities reappear after long temporal gaps. We present VideoMemory, an entity-centric framework that integrates narrative planning with visual generation through a Dynamic Memory Bank. Given a structured script, a multi-agent system decomposes the narrative into shots, retrieves entity representations from memory, and synthesizes keyframes and videos conditioned on these retrieved states. The Dynamic Memory Bank stores explicit visual and semantic descriptors for characters, props, and backgrounds, and is updated after each shot to reflect story-driven changes while preserving identity. This retrieval-update mechanism enables consistent portrayal of entities across distant shots and supports coherent long-form generation. To evaluate this setting, we construct a 54-case multi-shot consistency benchmark covering character-, prop-, and background-persistent scenarios. Extensive experiments show that VideoMemory achieves strong entity-level coherence and high perceptual quality across diverse narrative sequences.