ToonCrafter: Generative Cartoon Interpolation

TL;DR

ToonCrafter introduces a generative cartoon interpolation method that overcomes traditional limitations by adapting live-action priors, addressing complex motions and dis-occlusion, and providing user control for more natural and detailed results.

Contribution

The paper presents a novel framework that adapts live-action video priors to cartoons, introduces a dual-reference 3D decoder, and offers an interactive sketch encoder for improved cartoon video interpolation.

Findings

Produces more natural and convincing cartoon interpolations.

Effectively handles complex non-linear motions and dis-occlusion.

Outperforms existing methods in qualitative and quantitative evaluations.

Abstract

We introduce ToonCrafter, a novel approach that transcends traditional correspondence-based cartoon video interpolation, paving the way for generative interpolation. Traditional methods, that implicitly assume linear motion and the absence of complicated phenomena like dis-occlusion, often struggle with the exaggerated non-linear and large motions with occlusion commonly found in cartoons, resulting in implausible or even failed interpolation results. To overcome these limitations, we explore the potential of adapting live-action video priors to better suit cartoon interpolation within a generative framework. ToonCrafter effectively addresses the challenges faced when applying live-action video motion priors to generative cartoon interpolation. First, we design a toon rectification learning strategy that seamlessly adapts live-action video priors to the cartoon domain, resolving the domain gap and content leakage issues. Next, we introduce a dual-reference-based 3D decoder to compensate for lost details due to the highly compressed latent prior spaces, ensuring the preservation of fine details in interpolation results. Finally, we design a flexible sketch encoder that empowers users with interactive control over the interpolation results. Experimental results demonstrate that our proposed method not only produces visually convincing and more natural dynamics, but also effectively handles dis-occlusion. The comparative evaluation demonstrates the notable superiority of our approach over existing competitors.