Illumination-Aware Cross-Modality Differential Fusion Multispectral Pedestrian Detection

Multispectral information fusion technology is a practical approach to enhance pedestrian detection performance in low light conditions. However, current methods often overlook the impact of illumination on modal weights and significance inter-modal differential information. Therefore, this paper proposes novel illumination-aware cross-modality (IACMDF) model. The different modalities stage are adaptively adjusted according intensity scene. On other hand, advantages respective fully enhanced by amplifying suppressing commonality twin modalities. In addition, reduce loss problem caused importance occupied channels feature map convolutional pooling process, work adds squeeze-and-excitation attention mechanism after process. Experiments public multispectral dataset KAIST have shown that average miss rate our method substantially reduced compared baseline