Is Weakly-supervised Action Segmentation Ready For Human-Robot Interaction? No, Let's Improve It With Action-union Learning

Benchmark Model Rank Results
action-segmentation-on-50-salads-1AUL#19F1@50%: 67.1F1@25%: 81.3F1@10%: 84.4Acc: 77.9Edit: 77.0
action-segmentation-on-gtea-1AUL#19F1@50%: 67.3F1@25%: 85.5F1@10%: 88.2Acc: 69.2Edit: 84.0
weakly-supervised-action-localization-on-gteaAU-Action#1mAP@0.1:0.7: 76.9mAP@0.5: 66.3
weakly-supervised-action-segmentationAUL#1Acc: 67.3