Segment3D: Learning Fine-Grained Class-Agnostic 3D Segmentation without Manual Labels

TL;DR

Segment3D leverages 2D foundation models to generate automatic, high-quality 3D segmentation masks, enabling fine-grained, class-agnostic 3D scene segmentation without manual labels, thus improving generalization and scalability.

Contribution

It introduces a novel approach that uses 2D foundation models to automatically generate training labels for 3D segmentation, eliminating manual annotation.

Findings

Outperforms existing 3D segmentation models on fine-grained masks.

Enables easy addition of new training data without manual labels.

Improves generalization to unseen object classes.

Abstract

Current 3D scene segmentation methods are heavily dependent on manually annotated 3D training datasets. Such manual annotations are labor-intensive, and often lack fine-grained details. Importantly, models trained on this data typically struggle to recognize object classes beyond the annotated classes, i.e., they do not generalize well to unseen domains and require additional domain-specific annotations. In contrast, 2D foundation models demonstrate strong generalization and impressive zero-shot abilities, inspiring us to incorporate these characteristics from 2D models into 3D models. Therefore, we explore the use of image segmentation foundation models to automatically generate training labels for 3D segmentation. We propose Segment3D, a method for class-agnostic 3D scene segmentation that produces high-quality 3D segmentation masks. It improves over existing 3D segmentation models (especially on fine-grained masks), and enables easily adding new training data to further boost the segmentation performance -- all without the need for manual training labels.