Utonia: Toward One Encoder for All Point Clouds

TL;DR

Utonia introduces a unified self-supervised point transformer encoder trained across diverse 3D point cloud domains, enhancing perception and enabling benefits in robotics, multimodal reasoning, and spatial understanding.

Contribution

It is the first to train a single self-supervised encoder across multiple diverse point cloud domains, promoting a unified representation space.

Findings

Improves perception capabilities across domains.

Enhances robotic manipulation and multimodal reasoning.

Reveals emergent behaviors from joint training.

Abstract

We dream of a future where point clouds from all domains can come together to shape a single model that benefits them all. Toward this goal, we present Utonia, a first step toward training a single self-supervised point transformer encoder across diverse domains, spanning remote sensing, outdoor LiDAR, indoor RGB-D sequences, object-centric CAD models, and point clouds lifted from RGB-only videos. Despite their distinct sensing geometries, densities, and priors, Utonia learns a consistent representation space that transfers across domains. This unification improves perception capability while revealing intriguing emergent behaviors that arise only when domains are trained jointly. Beyond perception, we observe that Utonia representations can also benefit embodied and multimodal reasoning: conditioning vision-language-action policies on Utonia features improves robotic manipulation, and integrating them into vision-language models yields gains on spatial reasoning. We hope Utonia can serve as a step toward foundation models for sparse 3D data, and support downstream applications in AR/VR, robotics, and autonomous driving.