Exposure: A White-Box Photo Post-Processing Framework

TL;DR

This paper introduces a deep learning framework for photo retouching that learns from unpaired data, providing understandable edits and achieving results aligned with user preferences through reinforcement learning.

Contribution

It presents a novel white-box retouching system trained on unpaired data, combining deep CNNs with reinforcement learning for sequential, interpretable photo editing.

Findings

Achieves retouching results consistent with user preferences.

Provides an understandable, white-box editing process.

Outperforms baseline methods in user studies.

Abstract

Retouching can significantly elevate the visual appeal of photos, but many casual photographers lack the expertise to do this well. To address this problem, previous works have proposed automatic retouching systems based on supervised learning from paired training images acquired before and after manual editing. As it is difficult for users to acquire paired images that reflect their retouching preferences, we present in this paper a deep learning approach that is instead trained on unpaired data, namely a set of photographs that exhibits a retouching style the user likes, which is much easier to collect. Our system is formulated using deep convolutional neural networks that learn to apply different retouching operations on an input image. Network training with respect to various types of edits is enabled by modeling these retouching operations in a unified manner as resolution-independent differentiable filters. To apply the filters in a proper sequence and with suitable parameters, we employ a deep reinforcement learning approach that learns to make decisions on what action to take next, given the current state of the image. In contrast to many deep learning systems, ours provides users with an understandable solution in the form of conventional retouching edits, rather than just a "black-box" result. Through quantitative comparisons and user studies, we show that this technique generates retouching results consistent with the provided photo set.