IPCD: Intrinsic Point-Cloud Decomposition

TL;DR

The paper introduces IPCD, a novel method for decomposing colored point clouds into albedo and shade, enabling improved relighting and texture editing by addressing the challenges of non-grid structure and global-light estimation.

Contribution

It proposes IPCD-Net with point-wise feature aggregation and Projection-based Luminance Distribution for accurate intrinsic decomposition of point clouds.

Findings

IPCD-Net effectively reduces cast shadows in albedo.

Enhances color accuracy in shade decomposition.

Demonstrates applications in relighting and texture editing.

Abstract

Point clouds are widely used in various fields, including augmented reality (AR) and robotics, where relighting and texture editing are crucial for realistic visualization. Achieving these tasks requires accurately separating albedo from shade. However, performing this separation on point clouds presents two key challenges: (1) the non-grid structure of point clouds makes conventional image-based decomposition models ineffective, and (2) point-cloud models designed for other tasks do not explicitly consider global-light direction, resulting in inaccurate shade. In this paper, we introduce \textbf{Intrinsic Point-Cloud Decomposition (IPCD)}, which extends image decomposition to the direct decomposition of colored point clouds into albedo and shade. To overcome challenge (1), we propose \textbf{IPCD-Net} that extends image-based model with point-wise feature aggregation for non-grid data processing. For challenge (2), we introduce \textbf{Projection-based Luminance Distribution (PLD)} with a hierarchical feature refinement, capturing global-light ques via multi-view projection. For comprehensive evaluation, we create a synthetic outdoor-scene dataset. Experimental results demonstrate that IPCD-Net reduces cast shadows in albedo and enhances color accuracy in shade. Furthermore, we showcase its applications in texture editing, relighting, and point-cloud registration under varying illumination. Finally, we verify the real-world applicability of IPCD-Net.