The Aeroplane and Undercarriage Detection Based on Attention Mechanism and Multi-Scale Features Processing

Table 1

Performance comparison of six object detection networks.

Method	mAP (%)	Time (s)
Faster R-CNN (VGG16)	68.12	0.3756
YOLO V3	63.87	0.0486
YOLO V4	65.75	0.0463
YOLO V4 + Inception-ResNet block	67.67	0.0490
YOLO V4 + SE block	70.66	0.0475
Our algorithm	71.93	0.0502