Learning Data-driven Reflectance Priors for Intrinsic Image Decomposition

TL;DR

This paper introduces a data-driven method for intrinsic image decomposition that leverages learned reflectance priors from human annotations, improving results especially under challenging lighting conditions.

Contribution

It presents a novel two-stage learning framework that integrates a learned reflectance prior into energy minimization for better intrinsic image decomposition.

Findings

Outperforms state-of-the-art methods on decomposition tasks

Enhances image relighting accuracy in difficult lighting scenarios

Demonstrates the effectiveness of relative reflectance priors

Abstract

We propose a data-driven approach for intrinsic image decomposition, which is the process of inferring the confounding factors of reflectance and shading in an image. We pose this as a two-stage learning problem. First, we train a model to predict relative reflectance ordering between image patches (`brighter', `darker', `same') from large-scale human annotations, producing a data-driven reflectance prior. Second, we show how to naturally integrate this learned prior into existing energy minimization frameworks for intrinsic image decomposition. We compare our method to the state-of-the-art approach of Bell et al. on both decomposition and image relighting tasks, demonstrating the benefits of the simple relative reflectance prior, especially for scenes under challenging lighting conditions.