Publications

VideoFusion: A Spatio-Temporal Collaborative Network for Multi-modal Video Fusion and Restoration

DSPFusion: Image Fusion via Degradation and Semantic Dual-Prior Guidance

Mask-DiFuser: A Masked Diffusion Model for Unified Unsupervised Image Fusion

A masked diffusion model that unifies multiple unsupervised image-fusion tasks; now an ESI Highly Cited Paper (top 1%).

ControlFusion: A Controllable Image Fusion Framework with Language-Vision Degradation Prompts

C2RF: Bridging Multi-modal Image Registration and Fusion via Commonality Mining and Contrastive Learning

DRMF: Degradation-Robust Multi-Modal Image Fusion via Composable Diffusion Prior

Image Fusion in the Loop of High-level Vision Tasks: A Semantic-aware Real-time Infrared and Visible Image Fusion Network

Introduces semantic awareness and high-level task feedback into real-time infrared-visible fusion; an ESI Hot and Highly Cited Paper and the sole Information Fusion Best Paper Award winner for 2024.

SwinFusion: Cross-domain Long-range Learning for General Image Fusion via Swin Transformer

Models cross-domain long-range dependencies with a Swin Transformer for general image fusion; an ESI Hot and Highly Cited Paper and winner of the 2023 Hsue-shen Tsien Paper Award.

PIAFusion: A Progressive Infrared and Visible Image Fusion Network Based on Illumination Aware

Progressively fuses infrared and visible information according to illumination conditions for robust day-and-night performance; an ESI Hot and Highly Cited Paper.

SuperFusion: A Versatile Image Registration and Fusion Network with Semantic Awareness

Jointly addresses image registration and semantics-aware fusion for spatially misaligned multimodal inputs; an ESI Hot and Highly Cited Paper.

DIVFusion: Darkness-free Infrared and Visible Image Fusion

Suppresses dark regions and improves detail visibility in low-light infrared-visible fusion; an ESI Hot and Highly Cited Paper.

Rethinking the Necessity of Image Fusion in High-level Vision Tasks: A Practical Infrared and Visible Image Fusion Network Based on Progressive Semantic Injection and Scene Fidelity

Couples low-level fusion quality with high-level vision tasks through progressive semantic injection and scene fidelity; an ESI Hot and Highly Cited Paper.

STDFusionNet: An Infrared and Visible Image Fusion Network Based on Salient Target Detection

Uses salient-target detection to preserve critical thermal targets and structural information in fused images; an ESI Highly Cited Paper (top 1%).

Deep Learning-based Image Fusion: A Survey

DRLIE: Flexible Low-light Image Enhancement via Disentangled Representations

CAMF: An Interpretable Infrared and Visible Image Fusion Network Based on Class Activation Mapping

Infrared and Visible Image Fusion via Parallel Scene and Texture Learning