Unpaired Point Cloud Completion on Real Scans using Adversarial Training
This addresses the challenge of scan completion for real-world 3D scanning applications where paired data is unavailable, representing a novel approach in the field.
The paper tackles the problem of completing 3D scans without needing paired training data, enabling direct application to real-world scans, and demonstrates realistic completions on datasets like ScanNet and KITTI with quantitative evaluation on the 3D-EPN benchmark.
As 3D scanning solutions become increasingly popular, several deep learning setups have been developed geared towards that task of scan completion, i.e., plausibly filling in regions there were missed in the raw scans. These methods, however, largely rely on supervision in the form of paired training data, i.e., partial scans with corresponding desired completed scans. While these methods have been successfully demonstrated on synthetic data, the approaches cannot be directly used on real scans in absence of suitable paired training data. We develop a first approach that works directly on input point clouds, does not require paired training data, and hence can directly be applied to real scans for scan completion. We evaluate the approach qualitatively on several real-world datasets (ScanNet, Matterport, KITTI), quantitatively on 3D-EPN shape completion benchmark dataset, and demonstrate realistic completions under varying levels of incompleteness.