"Zero-Shot" Point Cloud Upsampling

TL;DR

This paper introduces ZSPU, a data-agnostic zero-shot point cloud upsampling method that leverages internal information of individual point clouds, reducing training time and improving generalization to unseen data.

Contribution

The paper proposes a novel zero-shot approach for point cloud upsampling that does not require external training data, enabling better generalization and faster training.

Findings

Achieves competitive or superior results on benchmark datasets.

Reduces training time significantly compared to supervised methods.

Demonstrates strong generalization to unseen point clouds.

Abstract

Recent supervised point cloud upsampling methods are restricted by the size of training data and are limited in terms of covering all object shapes. Besides the challenges faced due to data acquisition, the networks also struggle to generalize on unseen records. In this paper, we present an internal point cloud upsampling approach at a holistic level referred to as "Zero-Shot" Point Cloud Upsampling (ZSPU). Our approach is data agnostic and relies solely on the internal information provided by a particular point cloud without patching in both self-training and testing phases. This single-stream design significantly reduces the training time by learning the relation between low resolution (LR) point clouds and their high (original) resolution (HR) counterparts. This association will then provide super resolution (SR) outputs when original point clouds are loaded as input. ZSPU achieves competitive/superior quantitative and qualitative performances on benchmark datasets when compared with other upsampling methods.