CSCWD：面向边缘设备轻量级小目标检测的跨尺度通道级知识蒸馏
CSCWD: Cross-Scale Channel-wise Knowledge Distillation for Lightweight Tiny Object Detection on Edge Devices
- Islamic Revolution Comprehensive University（伊斯兰革命综合大学）
- K.N. Toosi University of Technology（KN图西理工大学）
机构由 AI 辅助整理，请以论文原文为准。
中文总结 AI 辅助
提出跨尺度通道级知识蒸馏（CSCWD），将教师高分辨率特征迁移至轻量学生模型，在不增加推理复杂度下提升小目标检测精度，实验验证了有效性。
中文摘要 AI 辅助
航空图像中的实时小目标检测受到极小目标弱空间证据和轻量级检测器高分辨率细节丢失的限制。本研究提出跨尺度通道级知识蒸馏（CSCWD），一种训练时框架，将高分辨率空间表示从YOLO11m-P2教师模型迁移至紧凑的YOLO11n学生模型，且不改变学生模型的推理架构。与传统的同尺度特征蒸馏不同，CSCWD在特征对齐后将教师P2层的监督迁移至学生P3层，同时在更深金字塔层保留同尺度蒸馏。在统一的七序列Drone-vs-Bird验证协议下，YOLO11n-CSCWD在交并比阈值为0.5时达到50.17%的平均精度（mAP@0.5）和59.73%的召回率，较匹配的CA-YOLO11n基线在mAP@0.5上提升2.92个百分点，召回率提升3.55个百分点。跨尺度对齐相较于相应的同尺度通道级蒸馏配置，进一步将mAP@0.5提升2.09个百分点。在DUT-Anti-UAV上的零样本评估中，mAP@0.5从48.29%提升至50.06%，无需目标域微调。包含该域是因为其具有挑战性的小目标使得低延迟、计算高效的检测尤为相关。在树莓派5上使用NCNN-FP16，分辨率为640x640时，具有258万参数的学生模型在82.32毫秒平均墙钟延迟（即12.15帧每秒）下达到50.32%的mAP@0.5，同时保持与匹配基线几乎相同的运行时和内存需求。实验结果支持跨尺度蒸馏在不增加推理时模型复杂度的情况下改进小目标检测。
英文摘要
Real-time tiny object detection in aerial imagery is constrained by the weak spatial evidence of very small objects and the loss of high-resolution detail in lightweight detectors. This study presents Cross-Scale Channel-wise Knowledge Distillation (CSCWD), a training-time framework that transfers high-resolution spatial representations from a YOLO11m-P2 teacher to a compact YOLO11n student without altering the student's inference architecture. Unlike conventional same-scale feature distillation, CSCWD transfers supervision from teacher P2 to student P3 after feature alignment while retaining same-scale distillation at deeper pyramid levels. Under the unified seven-sequence Drone-vs-Bird validation protocol, YOLO11n-CSCWD achieves 50.17% mean average precision at an intersection-over-union threshold of 0.5 (mAP@0.5) and 59.73% recall, improving the matched CA-YOLO11n baseline by 2.92 percentage points in mAP@0.5 and 3.55 points in recall. Cross-scale alignment further increases mAP@0.5 by 2.09 points over the corresponding same-scale channel-wise distillation configuration. In zero-shot evaluation on DUT-Anti-UAV, mAP@0.5 increases from 48.29% to 50.06% without target-domain fine-tuning. This domain was included because its challenging small targets make low-latency, computationally efficient detection particularly relevant. On Raspberry Pi 5 using NCNN-FP16 at 640x640 resolution, the 2.58-million-parameter student achieves 50.32% mAP@0.5 at 82.32 ms mean wall-clock latency, or 12.15 frames per second, while retaining essentially the same runtime and memory requirements as the matched baseline. The results support cross-scale distillation for improving tiny-target detection without increasing inference-time model complexity.
