UniFusion: A Unified Image Fusion Framework with Robust Representation and Source-Aware Preservation

2026-03-15

AI Summary

This work proposes the first unified framework for multimodal image fusion, addressing the limitation of existing methods that are typically confined to specific tasks and struggle to balance source fidelity and fusion quality in cross-task scenarios. The framework leverages DINOv3-derived semantic features to construct a shared semantic space, incorporates a reconstruction alignment loss to enhance source-aware consistency, and employs a two-level optimization strategy to jointly improve fusion performance. Extensive experiments demonstrate that the proposed method achieves superior visual quality, strong generalization across diverse fusion tasks—including multi-exposure, multi-focus, and multimodal image fusion—and robust adaptability in real-world settings.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Image fusion aims to integrate complementary information from multiple source images to produce a more informative and visually consistent representation, benefiting both human perception and downstream vision tasks. Despite recent progress, most existing fusion methods are designed for specific tasks (i.e., multi-modal, multi-exposure, or multi-focus fusion) and struggle to effectively preserve source information during the fusion process. This limitation primarily arises from task-specific architectures and the degradation of source information caused by deep-layer propagation. To overcome these issues, we propose UniFusion, a unified image fusion framework designed to achieve cross-task generalization. First, leveraging DINOv3 for modality-consistent feature extraction, UniFusion establishes a shared semantic space for diverse inputs. Second, to preserve the understanding of each source image, we introduce a reconstruction-alignment loss to maintain consistency between fused outputs and inputs. Finally, we employ a bilevel optimization strategy to decouple and jointly optimize reconstruction and fusion objectives, effectively balancing their coupling relationship and ensuring smooth convergence. Extensive experiments across multiple fusion tasks demonstrate UniFusion's superior visual quality, generalization ability, and adaptability to real-world scenarios. Code is available at https://github.com/dusongcheng/UniFusion.

Problem

• image fusion
• source information preservation
• cross-task generalization
• multi-modal fusion
• feature degradation

Innovation

• Unified Image Fusion
• DINOv3 Feature Extraction
• Reconstruction-Alignment Loss
• Bilevel Optimization
• Cross-Task Generalization