Verifier-Guided Synthetic Augmentation for 3D Human Shape Generation

2026-10-02

AI Summary

This study addresses the limitations of training data in 3D human generation, which often lead to conservative or anatomically disproportionate models. To overcome this, we propose a verifier-guided synthetic augmentation framework that combines global and modality-specific local PCA with elastic registration to generate candidate samples. These candidates are subsequently filtered through a fitting-free anatomical verifier before retraining the diffusion model, thereby balancing low-cost data augmentation with strict geometric constraints. Evaluated on the DFAUST dataset, our approach achieves a sample acceptance rate of 86.32% and a CP-AUC of 0.871. Furthermore, the proposed EAUC metric demonstrates a 32% improvement over baselines, significantly enhancing both the diversity and geometric plausibility of the generated 3D human bodies.

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.

Problem

• 3D human shape generation
• training data diversity
• generative modeling
• body proportions

Innovation

• Verifier-Guided Augmentation
• Mode-Local PCA
• Elastic Registration
• Diffusion Model
• 3D Human Shape Generation