Verifier-guided method improves 3D human shape generation diversity
Verifier-Guided Synthetic Augmentation for 3D Human Shape Generation
Computer Vision and Pattern Recognition
Summary
Generating realistic 3D human body shapes is difficult because models tend to produce either very similar shapes or unrealistic ones that don’t follow natural body proportions. The authors developed a system that creates many candidate shapes, then uses a special checker to keep only those with correct body proportions and details. They retrain a model with these verified shapes, making it better at producing diverse but still realistic human bodies. Their tests show this approach creates more varied yet believable 3D humans compared to other methods.
What this means in practice
- •For 3d character artists: Generate a wider variety of realistic human body models for digital characters while ensuring anatomically plausible proportions.
- •For virtual try-on developers: Improve the diversity and realism of 3D body models used in virtual clothing fitting systems by augmenting datasets with verified synthetic shapes.$Commercial implications: Enables creation of realistic personalized avatars for shoppers, enhancing virtual try-on experiences and fashion retail services.
Authors
Yuexuan Wu, Yang Xiang, Hamid Laga, Dip Das, Anuj Srivastava, Zhengwu Zhang
Abstract
Limited training data diversity constrains generative modeling of 3D human bodies: conservative models remain close to observed examples, whereas exploratory models often violate basic body proportions. We introduce a verifier-guided augmentation framework that uses global and mode-local PCA to generate inexpensive candidates, screens them using correspondence-derived skeletal proportions and body-part geometry, and retrains a diffusion model on accepted candidates. Elastic registration provides both the modal structure used by distributed PCA and the dense anatomical correspondence needed for scalable screening without per-candidate body-model fitting. A blinded human study supports the verifier as a conservative gatekeeper, favoring verifier-accepted over rejected outputs. We evaluate full-pool verifier acceptance separately from the coverage and departure of accepted samples and combine them through EAUC. On 4,498 registered DFAUST surfaces, distributed-PCA augmentation achieves 86.32% acceptance, the highest CP-AUC (0.871), and the highest EAUC (0.752), improving EAUC by 32% over real-only and self-augmented diffusion. These results show that mode-local, verifier-guided proposals broaden diffusion generation while maintaining high agreement with calibrated body measurements.