Skip to main content

Table 1 Comparison between R-CNNs based on various mAP

From: Detection of Martian dust storms using mask regional convolutional neural networks

Networks \(mAP_{25}\) \(mAP_{50}\) \(mAP_{75}\) Time (ms)
strategy 1 strategy 2 strategy 1 strategy 2 strategy 1 strategy 2 strategy 1 strategy 2
Fast R-CNN 64.1 64.8 59.1 58.1 32.2 32.1 350 320
Mask R-CNN 67.8 68.1 61.2 60.5 38.1 33.1 290 300
SPPnet 62.5 63.2 50.1 53.1 22.5 20.1 370 360
Current R-CNN 68.2 67.5 62.1 61.2 39.0 37.7 320 310
  1. The highest mAP and the shortest time are presented in bold font. The \(mAP_{25}\) \(mAP_{50}\), \(mAP_{75}\) are mAPs with \(th_{IoU}\) of \(25\%\), \(50\%\) and \(75\%\)