Abstract:Recently, learning-based multi-exposure fusion (MEF) methods have made significant improvements. However, these methods mainly focus on static scenes and are prone to generate ghosting artifacts when tackling a more common scenario, i.e., the input images include motion, due to the lack of a benchmark dataset and solution for dynamic scenes. In this paper, we fill this gap by creating an MEF dataset of dynamic scenes, which contains multi-exposure image sequences and their corresponding high-quality reference images. To construct such a dataset, we propose a 'static-for-dynamic' strategy to obtain multi-exposure sequences with motions and their corresponding reference images. To the best of our knowledge, this is the first MEF dataset of dynamic scenes. Correspondingly, we propose a deep dynamic MEF (DDMEF) framework to reconstruct a ghost-free high-quality image from only two differently exposed images of a dynamic scene. DDMEF is achieved through two steps: pre-enhancement-based alignment and privilege-information-guided fusion. The former pre-enhances the input images before alignment, which helps to address the misalignments caused by the significant exposure difference. The latter introduces a privilege distillation scheme with an information attention transfer loss, which effectively improves the deghosting ability of the fusion network. Extensive qualitative and quantitative experimental results show that the proposed method outperforms state-of-the-art dynamic MEF methods. The source code and dataset are released at https://github.com/Tx000/Deep_dynamicMEF.

Robust ghost-free multiexposure fusion for dynamic scenes.

Ghost Detection And Removal Based On Super-Pixel Grouping In Exposure Fusion

Detail Preserving Exposure Fusion for A Dual Sensor Camera

Fast detail-preserving exposure fusion

Multi-exposure image fusion based on well-exposedness assessment

Robust Multi-Exposure Image Fusion: A Structural Patch Decomposition Approach

A Precise Multi-Exposure Image Fusion Method Based on Low-level Features.

A Novel Deghosting Method for Exposure Fusion.

Multi-exposure electric power monitoring image fusion method without ghosting based on exposure fusion framework and color dissimilarity feature

Motion Area Based Exposure Fusion Algorithm for Ghost Removal in High Dynamic Range Video Generation

Low-light Image Pairs Fusion Method Based on Patch-match

Robust Ghost-Free High-Dynamic-Range Imaging by Visual Salience Based Bilateral Motion Detection and Stack Extension Based Exposure Fusion.

Deep Exposure Fusion with Deghosting via Homography Estimation and Attention Learning

Deep Multi-Exposure Image Fusion for Dynamic Scenes

Robust Fusion of Two Large-Exposure-Ratio Images

Ghost Removal Via Channel Attention in Exposure Fusion.

Visual Salience And Stack Extension Based Ghost Removal For High-Dynamic-Range Imaging

Perceptual Evaluation for Multi-Exposure Image Fusion of Dynamic Scenes

Ghost-free multi exposure image fusion technique using dense SIFT descriptor and guided filter

A ghost-free multi-exposure image fusion using adaptive alignment for static and dynamic images