MoZoo:Unleashing Video Diffusion power in animal fur and muscle simulation

Hugging Face Daily Papers Papers
animal-simulation video-diffusion fur-simulation muscle-simulation generative-model attention-mechanism synthetic-to-real

Summary

MoZoo is a generative diffusion model that synthesizes high-fidelity animal videos from coarse meshes, using novel attention mechanisms and a synthetic-to-real data pipeline.

The creation of cinematic-quality animal effects necessitates the precise modeling of muscle and fur dynamics, a process that remains both labor-intensive and computationally expensive within traditional production workflows. While generative diffusion models have shown promise in diverse artistic workflows, their capacity for high-fidelity animal simulation remains largely unexploited. We present MoZoo, a generative dynamics solver that bypasses conventional refinement to synthesize high-fidelity animal videos from coarse meshes under multimodal guidance. We propose Role-Aware RoPE (RAR-RoPE) which employs role-based index remapping to synchronize motion alignment while decoupling reference information via fixed temporal offsets. Complementing this, Asymmetric Decoupled Attention partitions the latent sequence to enforce a unidirectional information flow, effectively preventing feature interference and improving computational efficiency. To address the scarcity of high-quality training data, we introduce MoZoo-Data, a synthetic-to-real pipeline that leverages a rendering engine and an inverse mapping approach to construct a large-scale dataset of paired sequences. Furthermore, we establish MoZooBench, a comprehensive benchmark with 120 mesh-video pairs. Experimental results demonstrate that MoZoo achieves high-fidelity fur simulation across diverse animal skeletons and layouts, preserving superior temporal and structural consistency.
Original Article
View Cached Full Text

Cached at: 05/29/26, 03:00 AM

Paper page - MoZoo:Unleashing Video Diffusion power in animal fur and muscle simulation

Source: https://huggingface.co/papers/2605.13857

Abstract

MoZoo generates high-fidelity animal videos from coarse meshes using diffusion models with novel attention mechanisms and a synthetic-to-real data pipeline.

The creation of cinematic-quality animal effects necessitates the precise modeling of muscle and fur dynamics, a process that remains both labor-intensive and computationally expensive within traditional production workflows. Whilegenerative diffusion modelshave shown promise in diverse artistic workflows, their capacity for high-fidelity animal simulation remains largely unexploited. We present MoZoo, a generative dynamics solver that bypasses conventional refinement to synthesize high-fidelity animal videos from coarse meshes under multimodal guidance. We propose Role-Aware RoPE (RAR-RoPE) which employsrole-based index remappingto synchronizemotion alignmentwhile decoupling reference information via fixedtemporal offsets. Complementing this,Asymmetric Decoupled Attentionpartitions thelatent sequenceto enforce a unidirectional information flow, effectively preventingfeature interferenceand improving computational efficiency. To address the scarcity of high-quality training data, we introduce MoZoo-Data, asynthetic-to-real pipelinethat leverages arendering engineand aninverse mapping approachto construct a large-scale dataset of paired sequences. Furthermore, we establish MoZooBench, acomprehensive benchmarkwith 120mesh-video pairs. Experimental results demonstrate that MoZoo achieves high-fidelityfur simulationacross diverse animal skeletons and layouts, preserving superior temporal andstructural consistency.

View arXiv pageView PDFProject pageGitHub85Add to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.13857 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.13857 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.13857 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

robbyant/lingbot-video-moe-30b-a3b

Hugging Face Models Trending

LingBot-Video is the first open-source large-scale MoE video generation model for embodied intelligence, featuring efficient MoE architecture, massive embodied data training, and multi-reward system for high aesthetics, physical rationality, and task completion.

AnyMo: Scaling Any-Modality Conditional Motion Generation with Masked Modeling

Hugging Face Daily Papers

This paper introduces AnyMo, a unified multimodal framework for human motion generation that combines a Residual FSQ-based motion tokenizer with a scalable masked modeling transformer, along with the OmniHuMo dataset of over 5,000 hours of motion data to enable high-quality synthesis under arbitrary modality combinations.