Joint Embedding of 3D Scan and CAD Objects

TL;DR

This paper introduces a novel 3D CNN-based joint embedding space for aligning and retrieving CAD models from 3D scans, leveraging a new dataset and outperforming existing methods by 12%.

Contribution

It presents a new approach to learn a shared embedding space for scan and CAD geometry using a stacked hourglass network and introduces a dataset for fine-grained evaluation.

Findings

Outperforms state-of-the-art CAD retrieval by 12%.

Introduces a new dataset with ranked scan-CAD similarity annotations.

Develops a 3D CNN-based method with a stacked hourglass architecture.

Abstract

3D scan geometry and CAD models often contain complementary information towards understanding environments, which could be leveraged through establishing a mapping between the two domains. However, this is a challenging task due to strong, lower-level differences between scan and CAD geometry. We propose a novel approach to learn a joint embedding space between scan and CAD geometry, where semantically similar objects from both domains lie close together. To achieve this, we introduce a new 3D CNN-based approach to learn a joint embedding space representing object similarities across these domains. To learn a shared space where scan objects and CAD models can interlace, we propose a stacked hourglass approach to separate foreground and background from a scan object, and transform it to a complete, CAD-like representation to produce a shared embedding space. This embedding space can then be used for CAD model retrieval; to further enable this task, we introduce a new dataset of ranked scan-CAD similarity annotations, enabling new, fine-grained evaluation of CAD model retrieval to cluttered, noisy, partial scans. Our learned joint embedding outperforms current state of the art for CAD model retrieval by 12% in instance retrieval accuracy.