Action Recognition on NTU RGB+D 120 (One-shot protocol)
77.4AccuracyFiGMo
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| FiGMoInput Modality=2D Motion + Video2026.06 | 77.4 | — | — | — | — | — | |
| FiGMoInput Modality=2D Motion2026.06 | 69.2 | — | — | — | — | — | |
| MotionBERT (finetune)Training Mode=finetune, Evaluation Protocol=One-shot2022.10 | 67.4 | — | — | — | — | — | |
| MotionBERTInput Modality=2D skeleton detections2026.06 | 67.4 | — | — | — | — | — | |
| MotionBERT (scratch)Training Mode=scratch, Evaluation Protocol=One-shot2022.10 | 61 | — | — | — | — | — | |
| Skeleton-DML2020.12 | 54.2 | — | — | — | — | — | |
| Skeleton-DMLEvaluation Protocol=One-shot2022.10 | 54.2 | — | — | — | — | — | |
| Skeleton-DMLInput Modality=2D skeleton detections2026.06 | 54.2 | — | — | — | — | — | |
| SL-DMLProtocol=One-shot2020.04 | 50.9 | — | — | — | — | — | |
| SL-DML2020.12 | 50.9 | — | — | — | — | — | |
| SL-DMLEvaluation Protocol=One-shot2022.10 | 50.9 | — | — | — | — | — | |
| SL-DMLInput Modality=2D skeleton detections2026.06 | 50.9 | — | — | — | — | — | |
| TCN2020.12 | 46.5 | — | — | — | — | — | |
| TCN-based one-shot action recognitionone-shot=true, m=12021.02 | 46.5 | — | — | — | — | — | |
| TCN OneShotEvaluation Protocol=One-shot2022.10 | 46.5 | — | — | — | — | — | |
| TCN OneshotInput Modality=2D skeleton detections2026.06 | 46.5 | — | — | — | — | — | |
| APSRProtocol=One-shot2020.04 | 45.3 | — | — | — | — | — | |
| APSR2020.12 | 45.3 | — | — | — | — | — | |
| APSR2021.02 | 45.3 | — | — | — | — | — | |
| APSREvaluation Protocol=One-shot2022.10 | 45.3 | — | — | — | — | — | |
| APSRInput Modality=2D skeleton detections2026.06 | 45.3 | — | — | — | — | — | |
| Average PoolingProtocol=One-shot2020.04 | 42.9 | — | — | — | — | — | |
| Average Pooling2020.12 | 42.9 | — | — | — | — | — | |
| ST-LSTMpooling strategy=Average Pooling2021.02 | 42.9 | — | — | — | — | — | |
| ST-LSTM + AvgPoolEvaluation Protocol=One-shot2022.10 | 42.9 | — | — | — | — | — | |
| ST-LSTM + AvgPoolInput Modality=2D skeleton detections2026.06 | 42.9 | — | — | — | — | — | |
| Fully ConnectedProtocol=One-shot2020.04 | 42.1 | — | — | — | — | — | |
| Fully Connected2020.12 | 42.1 | — | — | — | — | — | |
| ST-LSTMconnection type=Fully Connected2021.02 | 42.1 | — | — | — | — | — | |
| ST-LSTM + FCEvaluation Protocol=One-shot2022.10 | 42.1 | — | — | — | — | — | |
| Attention NetworkProtocol=One-shot2020.04 | 41 | — | — | — | — | — | |
| Attention Network2020.12 | 41 | — | — | — | — | — | |
| ST-LSTMpooling strategy=Attention2021.02 | 41 | — | — | — | — | — | |
| ST-LSTM + AttentionEvaluation Protocol=One-shot2022.10 | 41 | — | — | — | — | — | |
| ALCA-GCNEvaluation Protocol=One-shot2025.04 | — | 38.7 | 46.6 | 51 | 53.7 | 57.6 | |
| APSREvaluation Protocol=One-shot2025.04 | — | 29.1 | 34.8 | 39.2 | 42.8 | 45.3 | |
| JEANIEEvaluation Protocol=One-shot2025.04 | — | 38.5 | 44.1 | 50.3 | 51.2 | 57 | |
| M&C-scaleEvaluation Protocol=One-shot2025.04 | — | 44.1 | 55.3 | 60.3 | 64.2 | 68.7 | |
| Skeleton-DMLEvaluation Protocol=One-shot2025.04 | — | 28.6 | 37.5 | 48.6 | 48 | 54.2 | |
| SkeletonXEvaluation Protocol=One-shot2025.04 | — | 48.2 | 54.9 | 61.6 | 65.6 | 69.1 | |
| SL-DMLEvaluation Protocol=One-shot2025.04 | — | 36.7 | 42.4 | 49 | 46.4 | 50.9 | |
| SMAMEvaluation Protocol=One-shot2025.04 | — | 35.8 | 46.2 | 51.7 | 52.2 | 56.4 | |
| STA-MLNEvaluation Protocol=One-shot2025.04 | — | 42.5 | 48.8 | 53.1 | 54.3 | 59.9 | |
| uDTWEvaluation Protocol=One-shot2025.04 | — | 32.2 | 39 | 41.2 | 45.3 | 49 |