MMA-UNet: A Multi-Modal Asymmetric UNet Architecture for Infrared and Visible Image Fusion
This work improves multi-modal image fusion for applications like surveillance or medical imaging by enhancing information retention and balance, though it is incremental as it builds on existing UNet architectures.
The paper tackled the problem of multi-modal image fusion by addressing symmetric fusion issues that cause information loss or bias, proposing MMA-UNet with specialized encoders and cross-scale fusion to achieve balanced fusion. It demonstrated superior performance over state-of-the-art methods in infrared and visible image fusion, producing visually natural and semantically rich results.
Multi-modal image fusion (MMIF) maps useful information from various modalities into the same representation space, thereby producing an informative fused image. However, the existing fusion algorithms tend to symmetrically fuse the multi-modal images, causing the loss of shallow information or bias towards a single modality in certain regions of the fusion results. In this study, we analyzed the spatial distribution differences of information in different modalities and proved that encoding features within the same network is not conducive to achieving simultaneous deep feature space alignment for multi-modal images. To overcome this issue, a Multi-Modal Asymmetric UNet (MMA-UNet) was proposed. We separately trained specialized feature encoders for different modal and implemented a cross-scale fusion strategy to maintain the features from different modalities within the same representation space, ensuring a balanced information fusion process. Furthermore, extensive fusion and downstream task experiments were conducted to demonstrate the efficiency of MMA-UNet in fusing infrared and visible image information, producing visually natural and semantically rich fusion results. Its performance surpasses that of the state-of-the-art comparison fusion methods.