3D Generative Model Latent Disentanglement via Local Eigenprojection
This work addresses the challenge of fine-grained control in 3D shape generation for digital human creation, representing an incremental improvement over existing methods.
The paper tackles the problem of controlling local shape attributes in 3D generative models for digital humans by introducing a novel loss function based on spectral geometry, resulting in improved latent disentanglement while maintaining generation capabilities and comparable training times.
Designing realistic digital humans is extremely complex. Most data-driven generative models used to simplify the creation of their underlying geometric shape do not offer control over the generation of local shape attributes. In this paper, we overcome this limitation by introducing a novel loss function grounded in spectral geometry and applicable to different neural-network-based generative models of 3D head and body meshes. Encouraging the latent variables of mesh variational autoencoders (VAEs) or generative adversarial networks (GANs) to follow the local eigenprojections of identity attributes, we improve latent disentanglement and properly decouple the attribute creation. Experimental results show that our local eigenprojection disentangled (LED) models not only offer improved disentanglement with respect to the state-of-the-art, but also maintain good generation capabilities with training times comparable to the vanilla implementations of the models.