THUMOS14

Overview edit

Variants: THUMOS' 14, THUMOS 2014, THUMOS’14, THUMOS'14, THUMOS14

Associated Benchmarks

This dataset is used in 3 benchmarks:

• Temporal Action Localization - Metrics: Avg mAP (0.3:0.7)
• Weakly Supervised Action Localization - Metrics: avg-mAP (0.3-0.7), avg-mAP (0.1:0.7), avg-mAP (0.1-0.5)
• Action Recognition - Metrics: Accuracy

Recent Benchmark Submissions

Research Papers

Recent papers with results on this dataset:

• HR-Pro: Point-supervised Temporal Action Localization via Hierarchical Reliability Propagation (2023) -
• Actor-agnostic Multi-label Action Recognition with Multi-modal Query (2023) -
• Proposal-Based Multiple Instance Learning for Weakly-Supervised Temporal Action Localization (2023) -
• BasicTAD: an Astounding RGB-Only Baseline for Temporal Action Detection (2022) -
• Learning Action Completeness from Points for Weakly-supervised Temporal Action Localization (2021) -
• Action Unit Memory Network for Weakly Supervised Temporal Action Localization (2021) -
• ACM-Net: Action Context Modeling Network for Weakly-Supervised Temporal Action Localization (2021) -
• Adaptive Mutual Supervision for Weakly-Supervised Temporal Action Localization (2021) -
• CoLA: Weakly-Supervised Temporal Action Localization with Snippet Contrastive Learning (2021) -
• Point-Level Temporal Action Localization: Bridging Fully-supervised Proposals to Weakly-supervised Losses (2020) -
• Two-Stream Consensus Network for Weakly-Supervised Temporal Action Localization (2020) -
• Adversarial Background-Aware Loss for Weakly-supervised Temporal Activity Localization (2020) -
• Weakly-supervised Temporal Action Localization by Uncertainty Modeling (2020) -
• Weakly-Supervised Action Localization with Expectation-Maximization Multi-Instance Learning (2020) -