AnimaX: Animating the Inanimate in 3D with Joint Video-Pose Diffusion Models

TL;DR

AnimaX introduces a novel 3D animation framework that leverages video diffusion models and multi-view pose sequences to generate diverse, high-fidelity articulated mesh animations from textual prompts, overcoming limitations of traditional methods.

Contribution

It transfers video-based motion priors to 3D animation using joint video-pose diffusion conditioned on text and templates, supporting arbitrary skeletons and achieving state-of-the-art results.

Findings

State-of-the-art generalization on VBench

High motion fidelity and efficiency

Supports diverse articulated meshes

Abstract

We present AnimaX, a feed-forward 3D animation framework that bridges the motion priors of video diffusion models with the controllable structure of skeleton-based animation. Traditional motion synthesis methods are either restricted to fixed skeletal topologies or require costly optimization in high-dimensional deformation spaces. In contrast, AnimaX effectively transfers video-based motion knowledge to the 3D domain, supporting diverse articulated meshes with arbitrary skeletons. Our method represents 3D motion as multi-view, multi-frame 2D pose maps, and enables joint video-pose diffusion conditioned on template renderings and a textual motion prompt. We introduce shared positional encodings and modality-aware embeddings to ensure spatial-temporal alignment between video and pose sequences, effectively transferring video priors to motion generation task. The resulting multi-view pose sequences are triangulated into 3D joint positions and converted into mesh animation via inverse kinematics. Trained on a newly curated dataset of 160,000 rigged sequences, AnimaX achieves state-of-the-art results on VBench in generalization, motion fidelity, and efficiency, offering a scalable solution for category-agnostic 3D animation. Project page: \href{https://anima-x.github.io/}{https://anima-x.github.io/}.