TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Papers/Background Suppression Network for Weakly-supervised Tempo...

Background Suppression Network for Weakly-supervised Temporal Action Localization

Pilhyeon Lee, Youngjung Uh, Hyeran Byun

2019-11-22Weakly Supervised Action LocalizationAction LocalizationWeakly-supervised Temporal Action LocalizationTemporal Action Localization
PaperPDFCode(official)Code

Abstract

Weakly-supervised temporal action localization is a very challenging problem because frame-wise labels are not given in the training stage while the only hint is video-level labels: whether each video contains action frames of interest. Previous methods aggregate frame-level class scores to produce video-level prediction and learn from video-level action labels. This formulation does not fully model the problem in that background frames are forced to be misclassified as action classes to predict video-level labels accurately. In this paper, we design Background Suppression Network (BaS-Net) which introduces an auxiliary class for background and has a two-branch weight-sharing architecture with an asymmetrical training strategy. This enables BaS-Net to suppress activations from background frames to improve localization performance. Extensive experiments demonstrate the effectiveness of BaS-Net and its superiority over the state-of-the-art methods on the most popular benchmarks - THUMOS'14 and ActivityNet. Our code and the trained model are available at https://github.com/Pilhyeon/BaSNet-pytorch.

Results

TaskDatasetMetricValueModel
VideoTHUMOS 2014mAP@0.1:0.543.6BaS-Net
VideoTHUMOS 2014mAP@0.1:0.735.3BaS-Net
VideoTHUMOS 2014mAP@0.527BaS-Net
VideoTHUMOS’14mAP@0.527BasNet
VideoActivityNet-1.3mAP@0.534.5BaS-Net
VideoActivityNet-1.3mAP@0.5:0.9522.2BaS-Net
VideoActivityNet-1.2mAP@0.538.5BaS-Net
Temporal Action LocalizationTHUMOS 2014mAP@0.1:0.543.6BaS-Net
Temporal Action LocalizationTHUMOS 2014mAP@0.1:0.735.3BaS-Net
Temporal Action LocalizationTHUMOS 2014mAP@0.527BaS-Net
Temporal Action LocalizationTHUMOS’14mAP@0.527BasNet
Temporal Action LocalizationActivityNet-1.3mAP@0.534.5BaS-Net
Temporal Action LocalizationActivityNet-1.3mAP@0.5:0.9522.2BaS-Net
Temporal Action LocalizationActivityNet-1.2mAP@0.538.5BaS-Net
Zero-Shot LearningTHUMOS 2014mAP@0.1:0.543.6BaS-Net
Zero-Shot LearningTHUMOS 2014mAP@0.1:0.735.3BaS-Net
Zero-Shot LearningTHUMOS 2014mAP@0.527BaS-Net
Zero-Shot LearningTHUMOS’14mAP@0.527BasNet
Zero-Shot LearningActivityNet-1.3mAP@0.534.5BaS-Net
Zero-Shot LearningActivityNet-1.3mAP@0.5:0.9522.2BaS-Net
Zero-Shot LearningActivityNet-1.2mAP@0.538.5BaS-Net
Action LocalizationTHUMOS 2014mAP@0.1:0.543.6BaS-Net
Action LocalizationTHUMOS 2014mAP@0.1:0.735.3BaS-Net
Action LocalizationTHUMOS 2014mAP@0.527BaS-Net
Action LocalizationTHUMOS’14mAP@0.527BasNet
Action LocalizationActivityNet-1.3mAP@0.534.5BaS-Net
Action LocalizationActivityNet-1.3mAP@0.5:0.9522.2BaS-Net
Action LocalizationActivityNet-1.2mAP@0.538.5BaS-Net
Weakly Supervised Action LocalizationTHUMOS 2014mAP@0.1:0.543.6BaS-Net
Weakly Supervised Action LocalizationTHUMOS 2014mAP@0.1:0.735.3BaS-Net
Weakly Supervised Action LocalizationTHUMOS 2014mAP@0.527BaS-Net
Weakly Supervised Action LocalizationTHUMOS’14mAP@0.527BasNet
Weakly Supervised Action LocalizationActivityNet-1.3mAP@0.534.5BaS-Net
Weakly Supervised Action LocalizationActivityNet-1.3mAP@0.5:0.9522.2BaS-Net
Weakly Supervised Action LocalizationActivityNet-1.2mAP@0.538.5BaS-Net

Related Papers

DVFL-Net: A Lightweight Distilled Video Focal Modulation Network for Spatio-Temporal Action Recognition2025-07-16Including Semantic Information via Word Embeddings for Skeleton-based Action Recognition2025-06-23Zero-Shot Temporal Interaction Localization for Egocentric Videos2025-06-04A Review on Coarse to Fine-Grained Animal Action Recognition2025-06-01LLM-powered Query Expansion for Enhancing Boundary Prediction in Language-driven Action Localization2025-05-30CLIP-AE: CLIP-assisted Cross-view Audio-Visual Enhancement for Unsupervised Temporal Action Localization2025-05-29DeepConvContext: A Multi-Scale Approach to Timeseries Classification in Human Activity Recognition2025-05-27ProTAL: A Drag-and-Link Video Programming Framework for Temporal Action Localization2025-05-23