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

Table 1.

Node and hyperedge attribute definitions.

More »

Table 1 Expand

Fig 1.

Hyper-GNN message passing flow chart.

More »

Fig 1 Expand

Fig 2.

Feature map resolution enhancement and detail restoration.

More »

Fig 2 Expand

Fig 3.

Schematic diagram of dynamic inference path selection.

More »

Fig 3 Expand

Fig 4.

Examples of multimodal samples in the OWOD dataset.

More »

Fig 4 Expand

Table 2.

Comparison of Hyper-MDR with recent state-of-the-art methods on the OWOD dataset.

More »

Table 2 Expand

Table 3.

Impact of each module on overall performance.

More »

Table 3 Expand

Table 4.

Statistics of the meta-policy controller’s output actions in different scenarios.

More »

Table 4 Expand

Fig 5.

Policy adjustment latency distribution in different scenarios.

More »

Fig 5 Expand

Fig 6.

T-SNE visualization of cross-modal feature alignment.

More »

Fig 6 Expand

Fig 7.

Time series of attention entropy in modal fusion.

More »

Fig 7 Expand

Fig 8.

ROC curve for unknown class detection.

More »

Fig 8 Expand

Fig 9.

Pseudo-labeling quality and continuous learning’s ability to resist forgetting.

More »

Fig 9 Expand

Fig 10.

Detection results for different object types.

More »

Fig 10 Expand

Fig 11.

GFLOPs statistics and efficiency gains of dynamic inference.

More »

Fig 11 Expand