Point-Cloud Completion with Pretrained Text-to-image Diffusion Models

TL;DR

This paper introduces SDS-Complete, a novel method that leverages pre-trained text-to-image diffusion models to complete incomplete 3D point clouds by utilizing text semantics, outperforming existing methods especially on out-of-distribution objects.

Contribution

It presents a new approach that uses pre-trained text-to-image diffusion models and test-time optimization for point-cloud completion without requiring extensive 3D datasets.

Findings

Reduces Chamfer loss by 50% on average compared to current methods.

Effectively reconstructs objects absent from common datasets.

Works well on real-world depth sensor and LiDAR data.

Abstract

Point-cloud data collected in real-world applications are often incomplete. Data is typically missing due to objects being observed from partial viewpoints, which only capture a specific perspective or angle. Additionally, data can be incomplete due to occlusion and low-resolution sampling. Existing completion approaches rely on datasets of predefined objects to guide the completion of noisy and incomplete, point clouds. However, these approaches perform poorly when tested on Out-Of-Distribution (OOD) objects, that are poorly represented in the training dataset. Here we leverage recent advances in text-guided image generation, which lead to major breakthroughs in text-guided shape generation. We describe an approach called SDS-Complete that uses a pre-trained text-to-image diffusion model and leverages the text semantics of a given incomplete point cloud of an object, to obtain a complete surface representation. SDS-Complete can complete a variety of objects using test-time optimization without expensive collection of 3D information. We evaluate SDS Complete on incomplete scanned objects, captured by real-world depth sensors and LiDAR scanners. We find that it effectively reconstructs objects that are absent from common datasets, reducing Chamfer loss by 50% on average compared with current methods. Project page: https://sds-complete.github.io/