Structure-Aware Shape Synthesis

TL;DR

This paper introduces a structure-aware loss function for training 3D shape generative models, ensuring generated shapes are consistent with an underlying structural summary, leading to more plausible and realistic shapes.

Contribution

It presents a novel end-to-end training method that incorporates structural information into 3D shape synthesis, improving structural plausibility.

Findings

Enhanced structural consistency in generated shapes

Reduction in implausible and unrealistic shapes

Improved robustness across diverse observations

Abstract

We propose a new procedure to guide training of a data-driven shape generative model using a structure-aware loss function. Complex 3D shapes often can be summarized using a coarsely defined structure which is consistent and robust across variety of observations. However, existing synthesis techniques do not account for structure during training, and thus often generate implausible and structurally unrealistic shapes. During training, we enforce structural constraints in order to enforce consistency and structure across the entire manifold. We propose a novel methodology for training 3D generative models that incorporates structural information into an end-to-end training pipeline.