PointInfinity: Resolution-Invariant Point Diffusion Models

Zixuan Huang; Justin Johnson; Shoubhik Debnath; James M. Rehg; Chao-Yuan Wu

返回文献探索

Paper · arXiv 2404.03566

PointInfinity: Resolution-Invariant Point Diffusion Models

Zixuan Huang, Justin Johnson, Shoubhik Debnath, James M. Rehg, Chao-Yuan Wu

14 upvotesApril 4, 2024arXiv 预印本

AI 摘要

PointInfinity, a transformer-based point cloud diffusion model, efficiently generates high-resolution point clouds with state-of-the-art quality by using a fixed-size, resolution-invariant latent representation.

point cloud diffusion modelstransformer-based architectureresolution-invariant latent representationclassifier-free guidanceCO3Dpoint cloudshigh-resolutionstate-of-the-art quality

Abstract

We present PointInfinity, an efficient family of point cloud diffusion models. Our core idea is to use a transformer-based architecture with a fixed-size, resolution-invariant latent representation. This enables efficient training with low-resolution point clouds, while allowing high-resolution point clouds to be generated during inference. More importantly, we show that scaling the test-time resolution beyond the training resolution improves the fidelity of generated point clouds and surfaces. We analyze this phenomenon and draw a link to classifier-free guidance commonly used in diffusion models, demonstrating that both allow trading off fidelity and variability during inference. Experiments on CO3D show that PointInfinity can efficiently generate high-resolution point clouds (up to 131k points, 31 times more than Point-E) with state-of-the-art quality.