ExpoMotion: A Large-Scale Benchmark and A Householder Projection Network for Multi-Exposure Fusion

2026-07-03

AI Summary

Dynamic multi-exposure fusion is prone to motion-induced ghosting artifacts and lacks high-quality, large-scale benchmark datasets. To address these challenges, this work introduces ExpoMotion, the first dynamic multi-exposure benchmark comprising 1,738 sequences, and proposes the HOP network. HOP leverages Global Prior Illumination Alignment (GPIA) for initial exposure alignment and innovatively models ghosting artifacts as orthogonal perturbations on feature manifolds, which are effectively removed via Householder orthogonal projection. Evaluated on ExpoMotion, the proposed method significantly suppresses ghosting while preserving high-frequency details, demonstrating superior fusion quality and strong generalization capability.

Abstract

Multi-Exposure Fusion (MEF) effectively extends dynamic range, but practical deployment is hindered by motion-induced ghosting and the scarcity of high-quality dynamic benchmarks. Current benchmarks largely neglect dynamic scenes and lack reliable ground truth, making it difficult to handle the complexity of real-world motions. In response, we introduce ExpoMotion, a large-scale benchmark designed to evaluate deghosting capabilities. Comprising 1,738 sequences and 10,909 images across diverse environments, it covers a wide range of motions and provides high-fidelity GTs constructed through an expert-guided acquisition pipeline. To tackle the complex dynamics and extreme conditions captured in this benchmark, we propose the Householder Orthogonal Projection network (HOP), which revisits MEF deghosting from a mathematical perspective via Householder transformation, decoupling multi-frame alignment into exposure pre-alignment and ghost filtering. Specifically, the Global Priors Illumination Alignment (GPIA) module first rectifies drastic dynamic range discrepancies by utilizing global statistics for exposure harmonization. Regarding ghost removal, our Householder Orthogonal Attention (HOA) models artifacts as orthogonal perturbations. By employing a dynamic Householder reflector, HOA effectively projects ghosts out of the feature manifold while preserving high-frequency details. Experiments demonstrate that our ExpoMotion dataset enables superior generalization and artifact-free detail restoration, while also validating the effectiveness and efficiency of the HOP method. The dataset and code are available at https://github.com/Leo-LiuYao/ExpoMotion.

Problem

• Multi-Exposure Fusion
• ghosting
• dynamic scenes
• benchmark
• ground truth

Innovation

• Multi-Exposure Fusion
• Householder Transformation
• Deghosting
• Dynamic Benchmark
• Orthogonal Projection