LaS-Comp: Zero-shot 3D Completion with Latent-Spatial Consistency

TL;DR

LaS-Comp is a zero-shot, category-agnostic 3D shape completion method that uses 3D foundation models' priors, featuring a two-stage process for faithful and seamless completion, and is evaluated on a new comprehensive benchmark.

Contribution

It introduces a training-free, two-stage 3D completion framework leveraging foundation models and presents Omni-Comp, a new benchmark for realistic evaluation.

Findings

Outperforms previous state-of-the-art methods in 3D completion.

Effective across diverse partial observations and categories.

Demonstrates robustness on real-world and synthetic data.

Abstract

This paper introduces LaS-Comp, a zero-shot and category-agnostic approach that leverages the rich geometric priors of 3D foundation models to enable 3D shape completion across diverse types of partial observations. Our contributions are threefold: First, \ourname{} harnesses these powerful generative priors for completion through a complementary two-stage design: (i) an explicit replacement stage that preserves the partial observation geometry to ensure faithful completion; and (ii) an implicit refinement stage ensures seamless boundaries between the observed and synthesized regions. Second, our framework is training-free and compatible with different 3D foundation models. Third, we introduce Omni-Comp, a comprehensive benchmark combining real-world and synthetic data with diverse and challenging partial patterns, enabling a more thorough and realistic evaluation. Both quantitative and qualitative experiments demonstrate that our approach outperforms previous state-of-the-art approaches. Our code and data will be available at \href{https://github.com/DavidYan2001/LaS-Comp}{LaS-Comp}.