<p>As an emerging solution, multispectral object detection based on the fusion of visible and infrared images has attracted increasing attention. Benefiting from the complementary information between the two modalities, multispectral approaches typically outperform single-modality methods. However, achieving efficient cross-modal feature fusion remains a central challenge. To address this, we propose a novel multispectral object detection framework, termed the Mamba and Wavelet Enhanced Fusion for Multispectral Object Detection (MWEFDet). The framework consists of two key stages: intra-scale fusion and inter-scale fusion. The intra-scale stage employs an Adaptive Interaction Enhancement Module to facilitate effective cross-modal feature interaction, while the inter-scale stage integrates the wavelet transform with the Mamba architecture to enhance multi-scale representation. Extensive experiments on two public datasets demonstrate that MWEFDet achieves state-of-the-art performance, exhibiting strong competitiveness against current leading methods.</p>