3D Generative Model Latent Disentanglement via Local Eigenprojection

Abstract

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.

arXiv commentComputer Graphics Forum 2023

References in corpus (10)

• The Emerging Field of Signal Processing on Graphs: Extending High-Dimensional Data Analysis to Networks and Other Irregular Domains
• Learning from Synthetic Humans
• GANSpace: Discovering Interpretable GAN Controls
• STAR: Sparse Trained Articulated Human Body Regressor
• Learning Skeletal Articulations with Neural Blend Shapes
• Synthesizing Coupled 3D Face Modalities by Trunk-Branch Generative Adversarial Networks
• Fully Convolutional Mesh Autoencoder using Efficient Spatially Varying Kernels
• LIMP: Learning Latent Shape Representations with Metric Preservation Priors
• Intraoperative Liver Surface Completion with Graph Convolutional VAE
• Disentangling Geometric Deformation Spaces in Generative Latent Shape Models