Shape Completion using 3D-Encoder-Predictor CNNs and Shape Synthesis

TL;DR

This paper presents a novel method combining deep neural networks and shape synthesis to accurately complete partial 3D shapes, producing detailed high-resolution models that respect global structure.

Contribution

It introduces a 3D-Encoder-Predictor Network for initial shape completion and a patch-based synthesis method for fine detail, integrating data-driven prediction with shape synthesis.

Findings

Outperforms state-of-the-art completion methods

Effective on real-world and synthetic data

Produces high-resolution, detailed 3D shapes

Abstract

We introduce a data-driven approach to complete partial 3D shapes through a combination of volumetric deep neural networks and 3D shape synthesis. From a partially-scanned input shape, our method first infers a low-resolution -- but complete -- output. To this end, we introduce a 3D-Encoder-Predictor Network (3D-EPN) which is composed of 3D convolutional layers. The network is trained to predict and fill in missing data, and operates on an implicit surface representation that encodes both known and unknown space. This allows us to predict global structure in unknown areas at high accuracy. We then correlate these intermediary results with 3D geometry from a shape database at test time. In a final pass, we propose a patch-based 3D shape synthesis method that imposes the 3D geometry from these retrieved shapes as constraints on the coarsely-completed mesh. This synthesis process enables us to reconstruct fine-scale detail and generate high-resolution output while respecting the global mesh structure obtained by the 3D-EPN. Although our 3D-EPN outperforms state-of-the-art completion method, the main contribution in our work lies in the combination of a data-driven shape predictor and analytic 3D shape synthesis. In our results, we show extensive evaluations on a newly-introduced shape completion benchmark for both real-world and synthetic data.