FlowRig: Unified Skeleton Generation for 3D and 4D Meshes via Flow Matching
cs.CV, cs.GR
Submitted: 2026-11-02
Updated: 2026-11-04
Comments: Project page: https://tamu-visual-ai.github.io/FlowRig/
Project page: https://tamu-visual-ai.github.io/FlowRig/ABSTRACT
License: http://creativecommons.org/licenses/by/4.0/
The gist: Automatic rigging is a fundamental task for both static 3D and dynamic 4D mesh animation, yet existing methods do not provide a unified feedforward solution for these two settings.
Terminology
Abstract
Automatic rigging is a fundamental task for both static 3D and dynamic 4D mesh animation, yet existing methods do not provide a unified feedforward solution for these two settings. While static 3D riggers achieve strong quality, they often provide limited control over bone count and become temporally unstable when applied frame by frame to dynamic meshes. Existing 4D methods, in contrast, typically rely on expensive per-sequence optimization. We propose FlowRig, a template-free, data-efficient framework for unified 3D and 4D skeleton generation. Built on flow matching, FlowRig generates temporally aligned skeletons for an entire sequence in a single feedforward pass, while naturally handling single-mesh 3D rigging. The model further supports explicit skeleton-count control, uncertainty estimation for candidate selection, and masking-based interpolation, forecasting, and backcasting. To address limited paired 4D rigging data, we adopt a curriculum that initializes from large-scale 3D rigging data and adapts to mid-scale 4D mesh-skeleton data. Experiments show that FlowRig achieves strong static 3D rigging performance with controllable bone complexity, as well as high-quality and temporally consistent 4D reconstruction and prediction results.
Sources
- CT4D: Consistent Text-to-4D Generation with Animatable Meshes
- Sync4D: Video Guided Controllable Dynamics for Physics-Based 4D Generation
- GaussianFlow: Splatting Gaussian Dynamics for 4D Content Creation
- AniGen: Unified $S^3$ Fields for Animatable 3D Asset Generation
- Diffusion4D: Fast Spatial-temporal Consistent 4D Generation via Video Diffusion Models
- Muses: Designing, Composing, Generating Nonexistent Fantasy 3D Creatures without Training
- PLA4D: Pixel-Level Alignments for Text-to-4D Gaussian Splatting
- SMF: Template-free and Rig-free Animation Transfer using Kinetic Codes
- DreamGaussian4D: Generative 4D Gaussian Splatting
- ActionMesh: Animated 3D Mesh Generation with Temporal 3D Diffusion
- GaussiAnimate: Reconstruct and Rig Animatable Categories with Level of Dynamics
- SPRig: Self-Supervised Pose-Invariant Rigging from Mesh Sequences
- SkeletonGaussian: Editable 4D Generation through Gaussian Skeletonization
- RigNet: Neural Rigging for Articulated Characters
- 4DGen: Grounded 4D Content Generation with Spatial-temporal Consistency
- 4Dynamic: Text-to-4D Generation with Hybrid Priors
- Trans4D: Realistic Geometry-Aware Transition for Compositional Text-to-4D Synthesis
- MagicPose4D: Crafting Articulated Models with Appearance and Motion Control
- RigMo: Unifying Rig and Motion Learning for Generative Animation
- Skin Tokens: A Learned Compact Representation for Unified Autoregressive Rigging
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs