A new framework enhances video world models’ long-term consistency by integrating a geometry-grounded long-term spatial memory mechanism.
Emerging world models autoregressively generate video frames in response to
actions, such as camera movements and text prompts, among other control
signals. Due to limited temporal context window sizes, these models often
struggle to maintain scene consistency during revisits, leading to severe
forgetting of previously generated environments. Inspired by the mechanisms of
human memory, we introduce a novel framework to enhancing long-term consistency
of video world models through a geometry-grounded long-term spatial memory. Our
framework includes mechanisms to store and retrieve information from the
long-term spatial memory and we curate custom datasets to train and evaluate
world models with explicitly stored 3D memory mechanisms. Our evaluations show
improved quality, consistency, and context length compared to relevant
baselines, paving the way towards long-term consistent world generation.
