Mobius: Text to Seamless Looping Video Generation via Latent Shift

TL;DR

Mobius is a new method that generates seamless, looping videos directly from text prompts using a pre-trained diffusion model, without additional training or user annotations, enabling dynamic and high-quality visual content.

Contribution

The paper introduces Mobius, a novel latent-shifting technique for text-to-looping video generation that does not require training or image-based appearance constraints.

Findings

Successfully generates seamless looping videos from text prompts.

Outperforms previous methods in visual quality and motion dynamics.

Flexible latent cycle length allows diverse looping effects.

Abstract

We present Mobius, a novel method to generate seamlessly looping videos from text descriptions directly without any user annotations, thereby creating new visual materials for the multi-media presentation. Our method repurposes the pre-trained video latent diffusion model for generating looping videos from text prompts without any training. During inference, we first construct a latent cycle by connecting the starting and ending noise of the videos. Given that the temporal consistency can be maintained by the context of the video diffusion model, we perform multi-frame latent denoising by gradually shifting the first-frame latent to the end in each step. As a result, the denoising context varies in each step while maintaining consistency throughout the inference process. Moreover, the latent cycle in our method can be of any length. This extends our latent-shifting approach to generate seamless looping videos beyond the scope of the video diffusion model's context. Unlike previous cinemagraphs, the proposed method does not require an image as appearance, which will restrict the motions of the generated results. Instead, our method can produce more dynamic motion and better visual quality. We conduct multiple experiments and comparisons to verify the effectiveness of the proposed method, demonstrating its efficacy in different scenarios. All the code will be made available.