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.

Main lines and fold characteristics of palmprint (left) and ROI (right).

More »

Fig 1 Expand

Fig 2.

Illustration of the method of extracting ROI by keypoints positioning in literature.

More »

Fig 2 Expand

Fig 3.

Flow structure of the proposed method.

More »

Fig 3 Expand

Fig 4.

Example images of SCAUPD.

More »

Fig 4 Expand

Fig 5.

Example images of the database used: (A) REST; (B) MPD; (C) IITD; (D) BMPD.

More »

Fig 5 Expand

Fig 6.

YOLOv5-lite network structure.

More »

Fig 6 Expand

Fig 7.

YOLOv5-lite palm initial localization output.

More »

Fig 7 Expand

Fig 8.

UNet network structure.

More »

Fig 8 Expand

Fig 9.

Proposed network structure.

More »

Fig 9 Expand

Fig 10.

Example of residual downsampling, input size (h, w, c), output size (h/2, w/2, c).

More »

Fig 10 Expand

Fig 11.

Depthwise separable convolution.

More »

Fig 11 Expand

Fig 12.

Keypoints in the original image (left) and keypoints visualized under the heatmap (right).

More »

Fig 12 Expand

Fig 13.

Schematic diagram of ROI extraction algorithm.

More »

Fig 13 Expand

Fig 14.

Example of data annotation.

More »

Fig 14 Expand

Table 1.

Statistics of the dataset used in the experiment.

More »

Table 1 Expand

Fig 15.

Training and validation loss curves for YOLOv5-lite.

More »

Fig 15 Expand

Fig 16.

Training and validation accuracy curves for YOLOv5-lite.

More »

Fig 16 Expand

Table 2.

Comparison of experimental results between YOLOv5-lite and other lightweight detection networks.

More »

Table 2 Expand

Fig 17.

Training and testing loss function curves for the improved UNet model.

More »

Fig 17 Expand

Fig 18.

Training and testing accuracy curves for the improved UNet model.

More »

Fig 18 Expand

Fig 19.

Threshold = 2 model output image (left) and normal label model output image (right).

More »

Fig 19 Expand

Table 3.

Model training accuracy and test accuracy under different thresholds.

More »

Table 3 Expand

Fig 20.

The implementation effect of the model on different datasets when Threshold = 1.5.

(A) is BMPD, (B) is IITD, (C) is MPD, (D) is REST, and (E) is SCAUPD.

More »

Fig 20 Expand

Fig 21.

Convergence curves of model accuracy under different loss functions.

More »

Fig 21 Expand

Table 4.

Comparison of test accuracies for the model under different loss functions.

More »

Table 4 Expand

Fig 22.

Convergence curve of accuracy for the residual block ablation experiment.

More »

Fig 22 Expand

Table 5.

Comparison of model size, average detection time, and testing accuracy using different modules.

More »

Table 5 Expand

Fig 23.

Diagram of changes in testing accuracy between the proposed network and other networks.

More »

Fig 23 Expand

Table 6.

Comparison between the proposed network and other networks.

More »

Table 6 Expand

Table 7.

Open set testing of the proposed method compared to other advanced methods.

More »

Table 7 Expand

Fig 24.

The performance of the proposed method on the RPG1K dataset.

More »

Fig 24 Expand

Fig 25.

The performance of the proposed method in the example of PKLNet [26] localization failure.

More »

Fig 25 Expand

Table 8.

The average detection time and model size of the proposed method compared to other advanced methods.

More »

Table 8 Expand