Learning A Locally Unified 3D Point Cloud for View Synthesis

TL;DR

This paper introduces a deep learning approach that constructs a locally unified 3D point cloud from sparse views to improve view synthesis quality, effectively filling holes and enhancing details in novel view rendering.

Contribution

It proposes a novel method for learning a locally unified 3D point cloud from source views, combining adaptive fusion and geometry-guided image restoration for superior view synthesis.

Findings

Improves PSNR by over 4 dB on benchmark datasets.

Achieves more accurate visual details in synthesized views.

Outperforms state-of-the-art view synthesis methods.

Abstract

In this paper, we explore the problem of 3D point cloud representation-based view synthesis from a set of sparse source views. To tackle this challenging problem, we propose a new deep learning-based view synthesis paradigm that learns a locally unified 3D point cloud from source views. Specifically, we first construct sub-point clouds by projecting source views to 3D space based on their depth maps. Then, we learn the locally unified 3D point cloud by adaptively fusing points at a local neighborhood defined on the union of the sub-point clouds. Besides, we also propose a 3D geometry-guided image restoration module to fill the holes and recover high-frequency details of the rendered novel views. Experimental results on three benchmark datasets demonstrate that our method can improve the average PSNR by more than 4 dB while preserving more accurate visual details, compared with state-of-the-art view synthesis methods.