Weakly Supervised Action Localization

21 benchmarks55 papers

In this task, the training data consists of videos with a list of activities in them without any temporal boundary annotations. However, while testing, given a video, the algorithm should recognize the activities in the video and also provide the start and end time.

Benchmarks

Weakly Supervised Action Localization on THUMOS 2014

mAP@0.5mAP@0.1:0.7mAP@0.1:0.5

Weakly Supervised Action Localization on ActivityNet-1.2

mAP@0.5Mean mAP

Weakly Supervised Action Localization on ActivityNet-1.3

mAP@0.5mAP@0.5:0.95

Weakly Supervised Action Localization on THUMOS’14

mAP@0.5avg-mAP (0.1-0.5)

Weakly Supervised Action Localization on THUMOS14

avg-mAP (0.3-0.7)avg-mAP (0.1:0.7)avg-mAP (0.1-0.5)

Weakly Supervised Action Localization on GTEA

mAP@0.1:0.7mAP@0.5

Weakly Supervised Action Localization on BEOID

mAP@0.1:0.7mAP@0.5

Weakly Supervised Action Localization on FineAction

mAPmAP IOU@0.5mAP IOU@0.75mAP IOU@0.95

Weakly Supervised Action Localization on THUMOS' 14

avg-mAP (0.1:0.7)