What is HumanRig
HumanRig is a 3D humanoid character automatic rigging research project developed by the Alibaba team. It addresses the problem that existing rigging technologies are limited in development due to the lack of high-quality datasets, and promotes the automation of 3D character animation production by providing a large-scale, high-quality dataset and an innovative automatic rigging framework. The HumanRig dataset contains 11,434 T-pose meshes, follows a unified skeleton topology, and has diverse head and body proportions, filling the gap in existing datasets in terms of scale, diversity, and skeleton consistency. The automatic rigging framework uses a Prior-Guided Skeleton Estimator (PGSE) and a Mesh-Skeleton Mutual Attention Network (MSMAN) to achieve coarse-to-fine 3D skeleton joint regression and skinning weight estimation, generating characters that can be used for animation production, with performance superior to existing methods.
Main Features of HumanRig
- Provides a large-scale high-quality dataset: HumanRig is the first large-scale dataset designed specifically for the task of automatic rigging of 3D humanoid characters, containing 11,434 high-quality AI-generated humanoid meshes. All models are presented in T-pose and strictly follow the industry-standard skeleton topology, and can be directly applied to mainstream animation engines. This dataset has significant advantages in scale, diversity, and skeleton consistency, covering a wide range of character types from real people to cartoon characters, and even anthropomorphic animals.
- Prior-Guided Skeleton Estimator (PGSE): Based on projecting 2D prior information into 3D space, it initializes a coarse skeleton, greatly reducing the complexity of the rigging task.
- U-shaped Point Transformer as the mesh encoder: It abandons reliance on edge information of 3D meshes, enhancing the robustness of rigging on complex meshes.
- Mesh-Skeleton Mutual Attention Network (MSMAN): By fusing mesh and skeleton features in a high-level semantic space, it achieves joint optimization of skeleton construction and skinning.
Technical Principles of HumanRig
- HumanRig Dataset Construction：
- Generate diverse 2D images: Through AI-driven 2D image generation technology, diverse T-pose character images are generated from text descriptions.
- Generate high-quality 3D meshes: Using tools such as InstantMesh and Unique3D, 2D images are converted into high-quality 3D meshes.
- Screening and optimization: From 17,268 initial meshes, 14,662 high-quality models were selected, and Mixamo semi-automated tools were used for skeleton rigging, ultimately forming 11,434 high-quality rigged models.
- Automatic Rigging Process：
- Skeleton initialization: Generate a coarse skeleton through the PGSE module.
- Feature extraction: Use an MLP-based skeleton encoder and a U-shaped Point Transformer to extract skeleton and mesh features respectively.
- Feature fusion and optimization: Through the MSMAN module, skeleton and mesh features are fused to achieve coarse-to-fine skeleton joint regression and skinning weight estimation.
- Generate animated characters: Finally, the optimized skeleton and skinning weights are combined to generate characters that can be used for animation production.
HumanRig Project Address
- Project official website：https://c8241998.github.io/HumanRig/
- arXiv technical paper：https://arxiv.org/pdf/2412.02317
Application Scenarios of HumanRig
- Game development: In game development, HumanRig's automatic rigging technology can greatly reduce the time and cost of character animation production, and performs excellently when handling complex character models (such as characters with complex clothing or accessories).
- Film and television production: In the film and television industry, HumanRig's automatic rigging technology can quickly generate high-quality character rigs, significantly improving production efficiency.
- Virtual reality (VR) and augmented reality (AR): In virtual reality and augmented reality applications, real-time interactive character animation is key to enhancing immersion. HumanRig's automatic rigging technology can provide real-time skeletal animation support for virtual characters, ensuring the naturalness and smoothness of character movements.
- 3D digital humans: Through automatic rigging technology, Amap can quickly generate personalized 3D digital humans, providing users with a more interactive and entertaining navigation experience.