Innovative Lightweight Detection for Airborne Remote Sensing: Integrating G-Shuffle and Dynamic Multi-Scale Pyramid Networks

To address the challenges of significant scale variations and difficulty in extracting discriminative features for small targets in airborne remote sensing object detection, while considering the constraints and efficiency requirements of onboard systems, this paper proposes a lightweight detection algorithm. First, a structural re-parameterization strategy is applied to optimize depthwise separable convolutions, simplifying complex structures during the inference stage, which improves inference speed and memory utilization. Second, the G-Shuffle module is designed to significantly enhance feature extraction efficiency and inter-channel information interaction, balancing computational complexity and detection accuracy. Lastly, a Dynamic Multi-Scale Pyramid Network (DMSPNet) is introduced, employing Pyramid Convolution (PyConv) to effectively extract and fuse multi-scale features. Additionally, a channel attention network is integrated to establish channel interconnections, expanding the receptive field and enhancing the extraction of fine-grained features at different scales, thereby improving the detection capability for targets of various sizes, especially small targets. Experimental results demonstrate that the proposed method achieves a mAP@0.5 of 0.552 and mAP@0.5:0.95 of 0.265 on the AI-TOD dataset, with a 2.7ms improvement in inference speed and a 14% reduction in model parameters. These results indicate superior performance across multiple metrics compared to existing state-of-the-art methods, proving the method's potential and practicality for airborne remote sensing object detection.