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.

Contrast learning.

More »

Fig 1 Expand

Fig 2.

Overall architecture diagram of the model.

More »

Fig 2 Expand

Table 1.

Analysis of different p0 values on DukeMTMCTOMarket1501 task.

More »

Table 1 Expand

Table 2.

Analysis of different p0 values on Market1501TODukeMTMC task.

More »

Table 2 Expand

Fig 3.

resnet50featuremaps.

More »

Fig 3 Expand

Table 3.

Dataset characteristics.

More »

Table 3 Expand

Table 4.

mAP Values of different functions.

More »

Table 4 Expand

Fig 4.

DukeMTMCTOMarket1501 loss ce.

More »

Fig 4 Expand

Fig 5.

Market1501TODukeMTMC loss ce.

More »

Fig 5 Expand

Fig 6.

DukeMTMCTOMarket1501 loss ce soft.

More »

Fig 6 Expand

Fig 7.

Market1501TODukeMTMC loss ce soft.

More »

Fig 7 Expand

Fig 8.

DukeMTMCTOMarket1501 loss contrast.

More »

Fig 8 Expand

Fig 9.

Market1501TODukeMTMC loss contrast.

More »

Fig 9 Expand

Fig 10.

DukeMTMCTOMarket1501 mAP.

More »

Fig 10 Expand

Fig 11.

Market1501TODukeMTMC mAP.

More »

Fig 11 Expand

Fig 12.

t-SNE Market1501 k = 5.

More »

Fig 12 Expand

Fig 13.

t-SNE Market1501 k = 10.

More »

Fig 13 Expand

Fig 14.

t-SNE DukeMTMC k = 5.

More »

Fig 14 Expand

Fig 15.

t-SNE DukeMTMC k = 10.

More »

Fig 15 Expand

Table 5.

Performance comparison of different methods.

More »

Table 5 Expand

Table 6.

Performance comparison of different methods on cross-domain person re-identification tasks (mAP, Rank-1, Rank-5, Rank-10).

More »

Table 6 Expand