# A Coarse-to-Fine Framework for Learned Color Enhancement with Non-Local Attention

## TL;DR

This paper introduces a coarse-to-fine neural network framework with non-local attention for automatic color enhancement, effectively balancing global style adjustment and local detail refinement.

## Contribution

It presents a novel two-stage CNN approach with non-local attention for improved color enhancement, combining global and local adjustments in a unified model.

## Key findings

- Outperforms existing methods on benchmark datasets
- Effectively captures long-range dependencies in images
- Improves local detail preservation while maintaining global style

## Abstract

Automatic color enhancement is aimed to adaptively adjust photos to expected styles and tones. For current learned methods in this field, global harmonious perception and local details are hard to be well-considered in a single model simultaneously. To address this problem, we propose a coarse-to-fine framework with non-local attention for color enhancement in this paper. Within our framework, we propose to divide enhancement process into channel-wise enhancement and pixel-wise refinement performed by two cascaded Convolutional Neural Networks (CNNs). In channel-wise enhancement, our model predicts a global linear mapping for RGB channels of input images to perform global style adjustment. In pixel-wise refinement, we learn a refining mapping using residual learning for local adjustment. Further, we adopt a non-local attention block to capture the long-range dependencies from global information for subsequent fine-grained local refinement. We evaluate our proposed framework on the commonly using benchmark and conduct sufficient experiments to demonstrate each technical component within it.