Action Segmentation on JIGSAWS
83.4AccuracyLC-SC-CRF
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| LC-SC-CRFInput Modality=Sensor-based2016.08 | 83.4 | 0.768 | |
| Bidir LSTMInput Modality=Sensor-based2016.08 | 83.3 | 0.811 | |
| TricorNetconvolution length=15, hidden states=642017.05 | 82.9 | 86.8 | |
| TricorNetvariant=low, convolution length=15, hidden states=642017.05 | 82.2 | 84.9 | |
| TCNInput Modality=Vision-based2016.08 | 81.4 | 0.831 | |
| TCN2017.05 | 81.4 | 83.1 | |
| LSTMInput Modality=Sensor-based2016.08 | 80.5 | 0.753 | |
| TCNInput Modality=Sensor-based2016.08 | 79.6 | 0.858 | |
| TricorNetvariant=high, convolution length=15, hidden states=642017.05 | 79.4 | 83.3 | |
| SD-SDLInput Modality=Sensor-based2016.08 | 78.6 | 0.833 | |
| ST-CNNInput Modality=Vision-based2016.08 | 77.7 | 0.68 | |
| ST-CNN2017.05 | 77.7 | 68 | |
| Seg-ST-CNNInput Modality=Vision-based2016.08 | 74.7 | 0.666 | |
| Spatial CNNInput Modality=Vision-based2016.08 | 74 | 0.377 | |
| Spatial CNN2017.05 | 74 | 37.7 | |
| MSM-CRFInput Modality=Vision-based2016.08 | 71.7 | — | |
| MSM-CRF2017.05 | 71.7 | — | |
| IDTInput Modality=Vision-based2016.08 | 53.9 | 0.085 | |
| VGGInput Modality=Vision-based2016.08 | 45.9 | 0.243 |