Filters








214 Hits in 7.1 sec

RGB Video Based Tennis Action Recognition Using a Deep Historical Long Short-Term Memory [article]

Jiaxin Cai, Xin Tang
2018 arXiv   pre-print
In this paper, we propose weighted Long Short-Term Memory adopted with convolutional neural network representations for three dimensional tennis shots recognition.  ...  Then, a weighted Long Short-Term Memory decoder is introduced to take the output state at time t and the historical embedding feature at time t-1 to generate feature vector using a score weighting scheme  ...  on CNN and deep historical Long Short-Term Memory networks for tennis action recognition .  ... 
arXiv:1808.00845v2 fatcat:7qvnb7oot5bhfoenjczk4qi2de

Convolutional Long Short-Term Memory Hybrid Networks for Skeletal Based Human Action Recognition

2020 VOLUME-8 ISSUE-10, AUGUST 2019, REGULAR ISSUE  
The objective is to develop a time series image representation of the skeletal action data and use it for recognition through a convolutional long short-term deep learning framework.  ...  Temporal decomposition is executed on long short term memory (LSTM) with data changes along x , y and z position vectors of the skeleton.  ...  Here, we propose an end to end trainable recurrent CNN with long short-term memory (LSTM) network to improve spatio temporal image recognition task on skeletal human action recognition.  ... 
doi:10.35940/ijitee.c8085.019320 fatcat:ht2xoopelvcptprokk4dc2ryci

Deep Learning for Domain-Specific Action Recognition in Tennis

Silvia Vinyes Mora, William J. Knottenbelt
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
Video-based action recognition in sports can significantly contribute to these advances.  ...  In order for action recognition to be useful in sports analytics a finer-grained action classification is needed.  ...  The resulting sequences of features are then fed to a deep neural network consisting of three stacked long short term memory units (LSTMs), a particular type of recurrent neural network (RNN).  ... 
doi:10.1109/cvprw.2017.27 dblp:conf/cvpr/MoraK17 fatcat:n67b5hpxgzdgrehq7ffjjwpiam

An Improved Attention-based Spatiotemporal-stream Model for Action Recognition in Videos

dan liu, yunfeng ji, mao ye, yan gan, Jianwei zhang
2020 IEEE Access  
INDEX TERMS Action recognition, spatiotemporal-stream, attention module, Ping-Pong action dataset. This work is licensed under a Creative Commons Attribution 4.0 License.  ...  In this paper, we propose an improved spatiotemporal attention model based on the two-stream structure to recognize the different actions in videos.  ...  Recurrent Neural Networks (RNN) and its variant Long Short-Term Memory (LSTM) [12] have been explored for sequential applications for decades.  ... 
doi:10.1109/access.2020.2983355 fatcat:xwrwcxfqjzffjanyhgjicljnby

A Novel Feature-Selection Method for Human Activity Recognition in Videos

Nadia Tweit, Muath A. Obaidat, Majdi Rawashdeh, Abdalraoof K. Bsoul, Mohammed GH. Al Zamil
2022 Electronics  
Human Activity Recognition (HAR) is the process of identifying human actions in a specific environment.  ...  This paper proposes a technique to select a set of representative features that can accurately recognize human activities from video streams, while minimizing the recognition time and memory.  ...  According to the results, the proposed method of feature reduction utilizes an acceptable amount of memory and has a positive effect in terms of classification time.  ... 
doi:10.3390/electronics11050732 fatcat:els3suj56nh5fcviwaetmgak7i

Dense ResNet based Human Action Recognition using Novel Trajectory Maps on 3D Skeletal Data

2019 VOLUME-8 ISSUE-10, AUGUST 2019, REGULAR ISSUE  
In this paper we propose to use skeletal trajectory maps for the detection of human actions. A new ResNet based algorithm named dense ResNet has been proposed to perform the classification task.  ...  The machine learning research community is presently working on human action/activity recognition issue in real-time videos, and facing several hundreds of confronts.  ...  Recurrent neural networks with a long short term memory network (RNN-LSTMs) [6] constitute the second group of methods.  ... 
doi:10.35940/ijitee.i8670.078919 fatcat:nsjsgjbsgzac7b3ugih3khx5jy

Attend It Again: Recurrent Attention Convolutional Neural Network for Action Recognition

Haodong Yang, Jun Zhang, Shuohao Li, Jun Lei, Shiqi Chen
2018 Applied Sciences  
Human action recognition in videos is an important task with a broad range of applications.  ...  "Attention-again" model is a variant from traditional attention model for recognizing human activities and is embedded in two long short-term memory (LSTM) layers.  ...  long-term dependencies.  ... 
doi:10.3390/app8030383 fatcat:vpt2uenqjvbbhdu3wuothrftre

3D Convolutional Networks for Action Recognition: Application to Sport Gesture Recognition [article]

Pierre-Etienne Martin, J Benois-Pineau, R Péteri, A Zemmari, J Morlier
2022 arXiv   pre-print
In the chapter we are interested in the classification of continuous video takes with repeatable actions, such as strokes of table tennis.  ...  Filmed in a free marker less ecological environment, these videos represent a challenge from both segmentation and classification point of view.  ...  Numerous works use models based on temporal networks such as Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) [UAM + 18].  ... 
arXiv:2204.08460v1 fatcat:7ks2orf4m5ge5ggmafbeghbrsu

Sport Action Recognition with Siamese Spatio-Temporal CNNs: Application to Table Tennis

Pierre-Etienne Martin, Jenny Benois-Pineau, Renaud Peteri, Julien Morlier
2018 2018 International Conference on Content-Based Multimedia Indexing (CBMI)  
Human action recognition in video is one of the key problems in visual data interpretation.  ...  Despite intensive research, the recognition of actions with low inter-class variability remains a challenge.  ...  ACKNOWLEDGMENT We would like to thank Alain Coupet, Ronan Le Merrer and Mathieu Dubos for their involvement in the acquisition and annotations of the table tennis dataset, and the reviewers for their constructive  ... 
doi:10.1109/cbmi.2018.8516488 dblp:conf/cbmi/MartinBPM18 fatcat:3l7rqo2n7vgrnppdaqkqp4fi74

Action Recognition Based on Joint Trajectory Maps with Convolutional Neural Networks [article]

Pichao Wang and Wanqing Li and Chuankun Li and Yonghong Hou
2016 arXiv   pre-print
Such an image-based representation enables us to fine-tune existing ConvNets models for the classification of skeleton sequences without training the networks afresh.  ...  The proposed method was evaluated on four public benchmark datasets, the large NTU RGB+D Dataset, MSRC-12 Kinect Gesture Dataset (MSRC-12), G3D Dataset and UTD Multimodal Human Action Dataset (UTD-MHAD  ...  A RNN is typically considered as memory cells, which are sensitive to both short as well as long term patterns.  ... 
arXiv:1612.09401v1 fatcat:bsuh6tiwnjdt7jhrjb6k3y35vi

Multi-view region-adaptive multi-temporal DMM and RGB action recognition

Mahmoud Al-Faris, John P. Chiverton, Yanyan Yang, David Ndzi
2020 Pattern Analysis and Applications  
Human action recognition remains an important yet challenging task. This work proposes a novel action recognition system.  ...  The region-adaptive weights, based on localised motion, accentuate and differentiate parts of actions possessing faster motion.  ...  This work also benefitted with credit provided by Google for using their Google Cloud.  ... 
doi:10.1007/s10044-020-00886-5 fatcat:or4aro7s5rfy5lcxcntwnhmc2i

RGB-D-based Human Motion Recognition with Deep Learning: A Survey [article]

Pichao Wang and Wanqing Li and Philip Ogunbona and Jun Wan and Sergio Escalera
2018 arXiv   pre-print
Specifically, deep learning methods based on the CNN and RNN architectures have been adopted for motion recognition using RGB-D data.  ...  As a survey focused on the application of deep learning to RGB-D-based motion recognition, we explicitly discuss the advantages and limitations of existing techniques.  ...  Instead of keeping a long-term memory of the entire body's motion in the cell, Shahroudy et al. [109] proposed a part-aware LSTM human action learning model (P-LSTM) wherein memory is split across part-based  ... 
arXiv:1711.08362v2 fatcat:cugugpqeffcshnwwto4z2aw4ti

A Discriminative Deep Model with Feature Fusion and Temporal Attention for Human Action Recognition

Jiahui Yu, Hongwei Gao, Wei Yang, Yueqiu Jiang, Weihong Chin, Naoyuki Kubota, Zhaojie Ju
2020 IEEE Access  
For long-term relations, we update the present memory state via the weight-controlled attention module that enables the memory cell to store better long-term features.  ...  We propose a novel discriminative deep model (D3D-LSTM) based on 3D-CNN and LSTM for both single-target and interaction action recognition to improve the spatiotemporal processing performance.  ...  This approach improves the recognition rate of long-term complex actions. 4) A new RGB-D dataset for action recognition, termed as Real-set, is designed and collected.  ... 
doi:10.1109/access.2020.2977856 fatcat:fu7lgrx4k5egzn7vadfmojlurq

A Novel Deep Learning based Automated Academic Activities Recognition in Cyber-Physical Systems

Muhammad Wasim, Imran Ahmed, Jamil Ahmad, Mohammad Mehedi Hassan
2021 IEEE Access  
For activity recognition in long videos, [32] propose a method by using a Bi-Directional Long Short Term Memory (LSTM) network cable for learning long sequences in videos.  ...  [34] use a contextual relationship-based deep learning framework for group activity recognition in videos.  ... 
doi:10.1109/access.2021.3073890 fatcat:xhjox7yxuzemlox2724diz6fz4

Skeleton based Activity Recognition by Fusing Part-wise Spatio-temporal and Attention Driven Residues [article]

Chhavi Dhiman, Dinesh Kumar Vishwakarma, Paras Aggarwal
2019 arXiv   pre-print
There exist a wide range of intra class variations of the same actions and inter class similarity among the actions, at the same time, which makes the action recognition in videos very challenging.  ...  part for action recognition by applying weighted late fusion mechanism.  ...  of the complete action, the defined part-wise action descriptors, are learnt using two consecutive layers of Long Short Term Memory (LSTM) units.  ... 
arXiv:1912.00576v1 fatcat:4pg77axdxbd43p6sa6lt6fmnoe
« Previous Showing results 1 — 15 out of 214 results