MFM-point: Multi-scale Flow Matching for Point Cloud Generation

TL;DR

MFM-Point introduces a multi-scale flow matching framework for point cloud generation that enhances quality and scalability while maintaining simplicity, outperforming existing point-based methods and rivaling representation-based approaches.

Contribution

The paper proposes a novel multi-scale flow matching framework with a structured downsampling and upsampling strategy for improved point cloud generation.

Findings

Achieves state-of-the-art performance among point-based methods.

Demonstrates strong results in multi-category and high-resolution tasks.

Enhances scalability without additional training or inference costs.

Abstract

In recent years, point cloud generation has gained significant attention in 3D generative modeling. Among existing approaches, point-based methods directly generate point clouds without relying on other representations such as latent features, meshes, or voxels. These methods offer low training cost and algorithmic simplicity, but often underperform compared to representation-based approaches. In this paper, we propose MFM-Point, a multi-scale Flow Matching framework for point cloud generation that substantially improves the scalability and performance of point-based methods while preserving their simplicity and efficiency. Our multi-scale generation algorithm adopts a coarse-to-fine generation paradigm, enhancing generation quality and scalability without incurring additional training or inference overhead. A key challenge in developing such a multi-scale framework lies in preserving the geometric structure of unordered point clouds while ensuring smooth and consistent distributional transitions across resolutions. To address this, we introduce a structured downsampling and upsampling strategy that preserves geometry and maintains alignment between coarse and fine resolutions. Our experimental results demonstrate that MFM-Point achieves best-in-class performance among point-based methods and challenges the best representation-based methods. In particular, MFM-point demonstrates strong results in multi-category and high-resolution generation tasks.