UniPre3D: Unified Pre-training of 3D Point Cloud Models with Cross-Modal Gaussian Splatting

TL;DR

UniPre3D introduces a versatile pre-training approach for 3D point cloud models that leverages Gaussian primitives and cross-modal features, achieving effective learning across various scales and architectures.

Contribution

It is the first unified pre-training method applicable to point clouds of any scale and architecture, integrating Gaussian splatting and 2D features for enhanced 3D representation learning.

Findings

Effective across object- and scene-level tasks

Compatible with diverse 3D model architectures

Improves 3D understanding performance

Abstract

The scale diversity of point cloud data presents significant challenges in developing unified representation learning techniques for 3D vision. Currently, there are few unified 3D models, and no existing pre-training method is equally effective for both object- and scene-level point clouds. In this paper, we introduce UniPre3D, the first unified pre-training method that can be seamlessly applied to point clouds of any scale and 3D models of any architecture. Our approach predicts Gaussian primitives as the pre-training task and employs differentiable Gaussian splatting to render images, enabling precise pixel-level supervision and end-to-end optimization. To further regulate the complexity of the pre-training task and direct the model's focus toward geometric structures, we integrate 2D features from pre-trained image models to incorporate well-established texture knowledge. We validate the universal effectiveness of our proposed method through extensive experiments across a variety of object- and scene-level tasks, using diverse point cloud models as backbones. Code is available at https://github.com/wangzy22/UniPre3D.