scholarly journals Weakly Supervised Temporal Action Localization Using Deep Metric Learning

2020 ◽  
Author(s):  
Ashraful Islam ◽  
Richard J. Radke
Keyword(s):  
Metric Learning ◽  
Action Localization ◽  
Deep Metric Learning ◽  
Weakly Supervised ◽  
Temporal Action

A deep metric learning approach for weakly supervised loan default prediction1

2021 ◽  
pp. 1-13
Author(s):  
Kai Zhuang ◽  
Sen Wu ◽  
Xiaonan Gao
Keyword(s):  
Default Risk ◽  
Metric Learning ◽  
Real Life ◽  
Risk Scores ◽  
Loan Default ◽  
Default Prediction ◽  
Heterogeneous Features ◽  
Deep Metric Learning ◽  
Weakly Supervised ◽  
Decision Boundaries

To deal with the systematic risk of financial institutions and the rapid increasing of loan applications, it is becoming extremely important to automatically predict the default probability of a loan. However, this task is non-trivial due to the insufficient default samples, hard decision boundaries and numerous heterogeneous features. To the best of our knowledge, existing related researches fail in handling these three difficulties simultaneously. In this paper, we propose a weakly supervised loan default prediction model WEAKLOAN that systematically solves all these challenges based on deep metric learning. WEAKLOAN is composed of three key modules which are used for encoding loan features, learning evaluation metrics and calculating default risk scores. By doing so, WEAKLOAN can not only extract the features of a loan itself, but also model the hidden relationships in loan pairs. Extensive experiments on real-life datasets show that WEAKLOAN significantly outperforms all compared baselines even though the default loans for training are limited.

Multi-Hierarchical Category Supervision for Weakly-Supervised Temporal Action Localization

2021 ◽  
pp. 1-1
Author(s):  
Guozhang Li ◽  
Jie Li ◽  
Nannan Wang ◽  
Xinpeng Ding ◽  
Zhifeng Li ◽  
...  
Keyword(s):  
Action Localization ◽  
Weakly Supervised ◽  
Temporal Action

scholarly journals Relational Prototypical Network for Weakly Supervised Temporal Action Localization

2020 ◽  
Vol 34 (07) ◽  
pp. 11053-11060
Author(s):  
Linjiang Huang ◽  
Yan Huang ◽  
Wanli Ouyang ◽  
Liang Wang
Keyword(s):  
State Of The Art ◽  
Previous Method ◽  
The State ◽  
Localization Method ◽  
Action Localization ◽  
Art Methods ◽  
Prototype Learning ◽  
Weakly Supervised ◽  
Temporal Action

In this paper, we propose a weakly supervised temporal action localization method on untrimmed videos based on prototypical networks. We observe two challenges posed by weakly supervision, namely action-background separation and action relation construction. Unlike the previous method, we propose to achieve action-background separation only by the original videos. To achieve this, a clustering loss is adopted to separate actions from backgrounds and learn intra-compact features, which helps in detecting complete action instances. Besides, a similarity weighting module is devised to further separate actions from backgrounds. To effectively identify actions, we propose to construct relations among actions for prototype learning. A GCN-based prototype embedding module is introduced to generate relational prototypes. Experiments on THUMOS14 and ActivityNet1.2 datasets show that our method outperforms the state-of-the-art methods.

scholarly journals Weakly Supervised Temporal Action Localization with Segment-Level Labels

2021 ◽  
pp. 42-54
Author(s):  
Xinpeng Ding ◽  
Nannan Wang ◽  
Jie Li ◽  
Xinbo Gao
Keyword(s):  
Action Localization ◽  
Weakly Supervised ◽  
Temporal Action

scholarly journals Self-Supervised Video Action Localization with Adversarial Temporal Transforms

2021 ◽  
Author(s):  
Guoqiang Gong ◽  
Liangfeng Zheng ◽  
Wenhao Jiang ◽  
Yadong Mu
Keyword(s):  
State Of The Art ◽  
Experimental Results ◽  
Policy Network ◽  
Video Classification ◽  
Time Warping ◽  
Action Localization ◽  
Temporal Boundary ◽  
Consistency Constraint ◽  
Weakly Supervised ◽  
Temporal Action

Weakly-supervised temporal action localization aims to locate intervals of action instances with only video-level action labels for training. However, the localization results generated from video classification networks are often not accurate due to the lack of temporal boundary annotation of actions. Our motivating insight is that the temporal boundary of action should be stably predicted under various temporal transforms. This inspires a self-supervised equivariant transform consistency constraint. We design a set of temporal transform operations, including naive temporal down-sampling to learnable attention-piloted time warping. In our model, a localization network aims to perform well under all transforms, and another policy network is designed to choose a temporal transform at each iteration that adversarially brings localization result inconsistent with the localization network's. Additionally, we devise a self-refine module to enhance the completeness of action intervals harnessing temporal and semantic contexts. Experimental results on THUMOS14 and ActivityNet demonstrate that our model consistently outperforms the state-of-the-art weakly-supervised temporal action localization methods.

scholarly journals Background Suppression Network for Weakly-Supervised Temporal Action Localization

2020 ◽  
Vol 34 (07) ◽  
pp. 11320-11327 ◽  
Author(s):  
Pilhyeon Lee ◽  
Youngjung Uh ◽  
Hyeran Byun
Keyword(s):  
State Of The Art ◽  
Background Suppression ◽  
Not Given ◽  
Training Strategy ◽  
Training Stage ◽  
Action Localization ◽  
Localization Performance ◽  
Weakly Supervised ◽  
Branch Weight ◽  
Temporal Action

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.

Sign in / Sign up

Export Citation Format

Share Document