Action Segmentation On Assembly101
评估指标
Edit
F1@10%
F1@25%
F1@50%
MoF
评测结果
各个模型在此基准测试上的表现结果
模型名称 | Edit | F1@10% | F1@25% | F1@50% | MoF | Paper Title | Repository |
|---|---|---|---|---|---|---|---|
| C2F-TCN | 32.4 | 33.3 | 29.0 | 21.3 | 39.2 | Coarse to Fine Multi-Resolution Temporal Convolutional Network | |
| UVAST | 31.5 | 32.1 | 28.3 | 20.8 | 37.4 | Unified Fully and Timestamp Supervised Temporal Action Segmentation via Sequence to Sequence Translation | |
| MS-TCN++ | 30.7 | 31.6 | 27.8 | 20.6 | 37.1 | MS-TCN++: Multi-Stage Temporal Convolutional Network for Action Segmentation | |
| ProTAS(Offline) | 29.2 | 28.7 | 24.4 | 17.5 | 34.8 | Progress-Aware Online Action Segmentation for Egocentric Procedural Task Videos | |
| ASFormer | 30.5 | 33.4 | 29.2 | 21.4 | 38.8 | ASFormer: Transformer for Action Segmentation | |
| LTContext | 30.4 | 33.9 | 30.0 | 22.6 | 41.2 | How Much Temporal Long-Term Context is Needed for Action Segmentation? | |
| ASQuery | 35.3 | 37.8 | 35.6 | 29.4 | 40.4 | ASQuery: A Query-based Model for Action Segmentation |
0 of 7 row(s) selected.