MSR-Net: Multi-Scale Relighting Network for One-to-One Relighting

TL;DR

This paper introduces MSR-Net, a multi-scale hierarchical network for efficient and high-quality image relighting, leveraging multi-scale feature aggregation and multi-step training to improve performance and robustness.

Contribution

The paper presents a novel multi-scale hierarchical network architecture and a multi-step training strategy for improved image relighting.

Findings

Achieves high-quality relighting with improved efficiency.

Multi-step training with different loss functions boosts performance.

Robustness demonstrated across various illumination settings.

Abstract

Deep image relighting allows photo enhancement by illumination-specific retouching without human effort and so it is getting much interest lately. Most of the existing popular methods available for relighting are run-time intensive and memory inefficient. Keeping these issues in mind, we propose the use of Stacked Deep Multi-Scale Hierarchical Network, which aggregates features from each image at different scales. Our solution is differentiable and robust for translating image illumination setting from input image to target image. Additionally, we have also shown that using a multi-step training approach to this problem with two different loss functions can significantly boost performance and can achieve a high quality reconstruction of a relighted image.