Skip to main content

Table 3 The value of bbox loss obtained by different schemes after training 500 epochs

From: Vehicle logo detection using an IoAverage loss on dataset VLD100K-61

Value/scheme

\({\mathcal{L}}_{IoU}\)

\({\mathcal{L}}_{GIoU}\)

\({\mathcal{L}}_{DIoU}\)

\({\mathcal{L}}_{CIoU}\)

\({\mathcal{L}}_{IoAverage}\)

BBox loss

0.01556

0.01602

0.01607

0.01577

0.009357

Relative reduction %

39.86%

41.59%

41.77%

41.67%

  1. Bold indicates the results obtained by calculating the BBox loss using the method proposed in this paper
  2. Relative reduction is calculated based on the bbox loss calculated by the IoAverage scheme