Home Console Docs News Papers Tutorials Datasets Wiki SOTA LLM Models GPU Leaderboard Events

English

Weakly Supervised Action Localization On 5

Metrics

avg-mAP (0.1-0.5)

avg-mAP (0.1:0.7)

avg-mAP (0.3-0.7)

Results

Performance results of various models on this benchmark

				Paper Title	Repository
HR-Pro	71.6	60.3	51.1	HR-Pro: Point-supervised Temporal Action Localization via Hierarchical Reliability Propagation
LACP	62.7	52.8	44.5	Learning Action Completeness from Points for Weakly-supervised Temporal Action Localization
P-MIL	57.4	47.0	38.0	Proposal-Based Multiple Instance Learning for Weakly-Supervised Temporal Action Localization
Ju et al.	55.6	44.8	35.4	Point-Level Temporal Action Localization: Bridging Fully-supervised Proposals to Weakly-supervised Losses	-
ACM-Net	53.2	42.6	33.4	ACM-Net: Action Context Modeling Network for Weakly-Supervised Temporal Action Localization
AUMN	52.1	41.5	32.4	Action Unit Memory Network for Weakly Supervised Temporal Action Localization	-
AMS	52.0	42.3	32.4	Adaptive Mutual Supervision for Weakly-Supervised Temporal Action Localization	-
Lee et al.	51.6	41.9	32.9	Weakly-supervised Temporal Action Localization by Uncertainty Modeling
CoLA	50.3	40.9	32.1	CoLA: Weakly-Supervised Temporal Action Localization with Snippet Contrastive Learning
TSCN	47.0	37.8	28.8	Two-Stream Consensus Network for Weakly-Supervised Temporal Action Localization	-
A2CL-PT	46.9	37.8	30.6	Adversarial Background-Aware Loss for Weakly-supervised Temporal Activity Localization
EM-ML	44.9	37.7	30.4	Weakly-Supervised Action Localization with Expectation-Maximization Multi-Instance Learning

0 of 12 row(s) selected.

Weakly Supervised Action Localization On 5 | SOTA | HyperAI