STAP: Spatial-Temporal Attention-Aware Pooling for Action Recognition

Nguyên, Tam; Zheng, Shuguo; Yan, Shuicheng

doi:10.1109/tcsvt.2014.2333151

Cited by 72 publications

(34 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To further improve activity recognition, recent works focused on exploiting context [10,16,50], which represent and harness information in both temporal and/or spatial neighborhood, or on attention [51], which learns an adaptive confidence score to leverage this surrounding information. In this realm, Caba Heilbron et al [10] develop a semantic context encoder that exploits evidence of objects and scenes within video segments to improve activity detection effectiveness and efficiency.…”

Section: Activity Recognitionmentioning

confidence: 99%

“…More recently, several works use temporal context to localize activities in videos [16] or to generate proposals [28]. Furthermore, Nguyen et al [51] present a pooling method that uses spatio-temporal attention for enhanced action recognition, while Pei et al [53] use temporal attention to gate neighboring observations in a RNN framework. Note that attention is also widely used in video captioning [34,44,48].…”

Section: Activity Recognitionmentioning

confidence: 99%

See 1 more Smart Citation

SoccerNet: A Scalable Dataset for Action Spotting in Soccer Videos

Giancola

Amine

Dghaily

et al. 2018

2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

160

218

View full text Add to dashboard Cite

In this paper, we introduce SoccerNet, a benchmark for action spotting in soccer videos. The dataset is composed of 500 complete soccer games from six main European leagues, covering three seasons from 2014 to 2017 and a total duration of 764 hours. A total of 6,637 temporal annotations are automatically parsed from online match reports at a one minute resolution for three main classes of events (Goal, Yellow/Red Card, and Substitution). As such, the dataset is easily scalable. These annotations are manually refined to a one second resolution by anchoring them at a single timestamp following well-defined soccer rules. With an average of one event every 6.9 minutes, this dataset focuses on the problem of localizing very sparse events within long videos. We define the task of spotting as finding the anchors of soccer events in a video. Making use of recent developments in the realm of generic action recognition and detection in video, we provide strong baselines for detecting soccer events. We show that our best model for classifying temporal segments of length one minute reaches a mean Average Precision (mAP) of 67.8%. For the spotting task, our baseline reaches an Average-mAP of 49.7% for tolerances δ ranging from 5 to 60 seconds. Our dataset and models are available at https://silviogiancola.github.io/SoccerNet.

show abstract

Section: Activity Recognitionmentioning

confidence: 99%

Section: Activity Recognitionmentioning

confidence: 99%

SoccerNet: A Scalable Dataset for Action Spotting in Soccer Videos

Giancola

Amine

Dghaily

et al. 2018

2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

160

218

View full text Add to dashboard Cite

show abstract

“…Since the current solution is specific for image parsing, we are also interested in generalizing the proposed method to other recognition tasks, such as image retrieval, and general k-NN classification applications. We also plan to leverage our work to video domain, i.e., action recognition [31] and human fixation prediction [32].…”

Section: Discussionmentioning

confidence: 99%

Adaptive Nonparametric Image Parsing

Nguyên

Sepúlveda

et al. 2015

IEEE Trans. Circuits Syst. Video Technol.

Self Cite

View full text Add to dashboard Cite

Abstract-In this paper, we present an adaptive nonparametric solution to the image parsing task, namely annotating each image pixel with its corresponding category label. For a given test image, first, a locality-aware retrieval set is extracted from the training data based on super-pixel matching similarities, which are augmented with feature extraction for better differentiation of local super-pixels. Then, the category of each super-pixel is initialized by the majority vote of the k-nearest-neighbor superpixels in the retrieval set. Instead of fixing k as in traditional non-parametric approaches, here we propose a novel adaptive nonparametric approach which determines the sample-specific k for each test image. In particular, k is adaptively set to be the number of the fewest nearest super-pixels which the images in the retrieval set can use to get the best category prediction. Finally, the initial super-pixel labels are further refined by contextual smoothing. Extensive experiments on challenging datasets demonstrate the superiority of the new solution over other stateof-the-art nonparametric solutions.Index Terms-image parsing, scene understanding, adaptive nonparametric method.

show abstract

“…There have been some recent attempts [5,7] to extend objectness to actionness. They often measure the actionness by fusing different feature channels such as space-time saliency [24], optical flow [7], body configuration [16] and deep learning features [9], sometimes with human input like eye fixation [23]. However, compared to objectness, actionness in videos is still not sufficiently explored due to the computational intensity in video space and the subtlety of human actions.…”

Section: Related Workmentioning

confidence: 99%

Improving Human Action Recognition by Non-action Classification

Wang

Hoai

2016

2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

In this paper we consider the task of recognizing human actions in realistic video where human actions are dominated by irrelevant factors. We first study the benefits of removing non-action video segments, which are the ones that do not portray any human action. We then learn a nonaction classifier and use it to down-weight irrelevant video segments. The non-action classifier is trained using ActionThread, a dataset with shot-level annotation for the occurrence or absence of a human action. The non-action classifier can be used to identify non-action shots with high precision and subsequently used to improve the performance of action recognition systems.

show abstract

STAP: Spatial-Temporal Attention-Aware Pooling for Action Recognition

Cited by 72 publications

References 37 publications

SoccerNet: A Scalable Dataset for Action Spotting in Soccer Videos

SoccerNet: A Scalable Dataset for Action Spotting in Soccer Videos

Adaptive Nonparametric Image Parsing

Improving Human Action Recognition by Non-action Classification

Contact Info

Product

Resources

About