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.

Multi-person dance tiered posture recognition with cross progressive multi-resolution representation integration.

More »

Fig 1 Expand

Fig 2.

Human skeleton keypoints model.

(a) 15 key points. (b) 18 key points. (c) 25 key points.

More »

Fig 2 Expand

Fig 3.

Multi-person posture recognition process.

(a) Top-down. (b) Bottom-up.

More »

Fig 3 Expand

Fig 4.

YOLOv8.

More »

Fig 4 Expand

Table 1.

Detailed parameter settings of used YOLOV8.

More »

Table 1 Expand

Fig 5.

HRNET model.

More »

Fig 5 Expand

Fig 6.

CPMRI module.

More »

Fig 6 Expand

Fig 7.

Geometrically-based keypoint segmentation.

More »

Fig 7 Expand

Table 2.

The results of the ablation experiments.

More »

Table 2 Expand

Fig 8.

The training curve of the different ablation models.

More »

Fig 8 Expand

Fig 9.

The testing curve of the different ablation models.

More »

Fig 9 Expand

Fig 10.

Visualized results in the MSCOCO2017 dataset.

More »

Fig 10 Expand

Table 3.

The results of the contarst experiments.

Underlined denotes SOTA top-down approaches, italicized denotes SOTA bottom-up methods, and bold represents the best result in each metric.

More »

Table 3 Expand

Fig 11.

Visualized single-person results in the custom dataset.

More »

Fig 11 Expand

Fig 12.

Visualized multi-person results in the custom dataset.

More »

Fig 12 Expand