AMFD: Distillation via Adaptive Multimodal Fusion for Multispectral Pedestrian Detection

Multispectral pedestrian detection has been shown to be effective in improving performance in complex illumination scenarios. However, prevalent double-stream networks in multispectral detection employ two separate feature extraction branches for multi-modal data, leading to nearly double the inference time compared to single-stream networks utilizing only one feature extraction branch. This increased inference time has hindered the widespread employment of multispectral pedestrian detection in embedded devices for autonomous systems. To efficiently compress multispectral object detection networks, we propose a novel distillation method, the Adaptive Modal Fusion Distillation (AMFD) framework. Unlike traditional distillation methods, the AMFD framework fully leverages the original modal features from the teacher network, thereby significantly enhancing the performance of the student network. Specifically, a Modal Extraction Alignment (MEA) module is utilized to derive learning weights for student networks, integrating focal and global attention mechanisms. This methodology enables the student network to acquire optimal fusion strategies independent from that of teacher network without necessitating an additional feature fusion module. Furthermore, we present the SMOD dataset, a well-aligned challenging multispectral dataset for detection. Extensive experiments on the challenging KAIST, LLVIP, SUNRGB-D and SMOD datasets are conducted to validate the effectiveness of AMFD. The results demonstrate that our method outperforms existing state-of-the-art methods in both reducing log-average Miss Rate and improving mean Average Precision. The code is available at https://github.com/bigD233/AMFD.git.