Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

Samples of the rail defect dataset with 8 categories.

From left to right, the first row contains saplling, dirt, unknown, and gap, and the second row contains squat, crush, scratch, and crack.

More »

Fig 1 Expand

Fig 2.

Structure of the proposed MBDA framework.

The input image is a detection sample of YOLO v5.

More »

Fig 2 Expand

Fig 3.

Image augmentation methods within IAB.

More »

Fig 3 Expand

Fig 4.

Feature augmentation methods within FAB.

More »

Fig 4 Expand

Table 1.

Detection results of different methods over the rail defect datasets.

More »

Table 1 Expand

Fig 5.

Detection performances and validation losses.

The first row from left to right are YOLOv5s, YOLOv5s6, and YOLOv5m. The second row from left to right are FasterRCNN R50, FasterRCNN R101, and MBDA.

More »

Fig 5 Expand

Fig 6.

Detection result.

From left to right are ground truth, MBDA, FasterRCNN, and YOLO.

More »

Fig 6 Expand

Table 2.

Different MBDA structures.

More »

Table 2 Expand

Fig 7.

Number of backbones ensembled and their detection precisions.

More »

Fig 7 Expand

Table 3.

Effectiveness of image augmentation operations.

N/A refers to taking no image augmentation method. OS and BS are short for object swap and layer swap, others refer to randomly selected image augmentation methods other than the two newly designed operations.

More »

Table 3 Expand

Table 4.

Effectiveness of feature augmentation operations.

LS, CS, SC are short for layer swap, channel swap, spot cover, respectively.

More »

Table 4 Expand

Table 5.

The impact of the combination of image augmentation and feature augmentation operations.

More »

Table 5 Expand