Ref-MEF: Reference-Guided Flexible Gated Image Reconstruction Network for Multi-Exposure Image Fusion

TL;DR

This paper introduces Ref-MEF, a flexible deep learning method for fusing multiple images with varying exposures into a single high-quality image.

Contribution

The novel contribution is the development of a reference-guided, flexible neural network for multi-exposure fusion that handles variable input sizes efficiently.

Findings

Ref-MEF uses a reference-guided exposure correction module with attention mechanisms to enhance feature extraction.

The method achieves superior visual quality and computational efficiency with increasing input images.

A refined loss function incorporating gradient fidelity improves image detail and dynamic range.

Abstract

Multi-exposure image fusion (MEF) is a computational approach that amalgamates multiple images, each captured at varying exposure levels, into a singular, high-quality image that faithfully encapsulates the visual information from all the contributing images. Deep learning-based MEF methodologies often confront obstacles due to the inherent inflexibilities of neural network structures, presenting difficulties in dynamically handling an unpredictable amount of exposure inputs. In response to this challenge, we introduce Ref-MEF, a method for color image multi-exposure fusion guided by a reference image designed to deal with an uncertain amount of inputs. We establish a reference-guided exposure correction (REC) module based on channel attention and spatial attention, which can correct input features and enhance pre-extraction features. The exposure-guided feature fusion (EGFF) module combines original image information and uses Gaussian filter weights for feature fusion while keeping the feature dimensions constant. The image reconstruction is completed through a gated context aggregation network (GCAN) and global residual learning GRL. Our refined loss function incorporates gradient fidelity, producing high dynamic range images that are rich in detail and demonstrate superior visual quality. In evaluation metrics focused on image features, our method exhibits significant superiority and leads in holistic assessments as well. It is worth emphasizing that as the number of input images increases, our algorithm exhibits notable computational efficiency.

Figures