51,754 Hits in 6.5 sec

One-Shot Learning for Real-Time Action Recognition [chapter]

Sean Ryan Fanello, Ilaria Gori, Giorgio Metta, Francesca Odone
2013 Lecture Notes in Computer Science  
The goal of the paper is to develop a one-shot real-time learning and recognition system for 3D actions.  ...  The main contribution of the paper is a real-time system for one-shot action modeling; moreover we highlight the effectiveness of sparse coding techniques to represent 3D actions.  ...  to reconcile with real-time action recognition.  ... 
doi:10.1007/978-3-642-38628-2_4 fatcat:4ehvgwazsrcflcp2deyoydypjq

One-shot action recognition in challenging therapy scenarios [article]

Alberto Sabater, Laura Santos, Jose Santos-Victor, Alexandre Bernardino, Luis Montesano, Ana C. Murillo
2021 arXiv   pre-print
This work presents a novel approach for one-shot action recognition in the wild that computes motion representations robust to variable kinematic conditions.  ...  Our approach is evaluated on the public NTU-120 one-shot action recognition benchmark, outperforming previous action recognition models.  ...  Experimental setup 4.1.1 Datasets The proposed method is designed for one-shot online action recognition in real scenarios.  ... 
arXiv:2102.08997v4 fatcat:s5vzjm3yivcljibtaxxiw6dbza

ProtoGAN: Towards Few Shot Learning for Action Recognition [article]

Sai Kumar Dwivedi, Vikram Gupta, Rahul Mitra, Shuaib Ahmed, Arjun Jain
2019 arXiv   pre-print
Few-shot learning (FSL) for action recognition is a challenging task of recognizing novel action categories which are represented by few instances in the training data.  ...  Our synthesized examples for a novel class are semantically similar to real examples belonging to that class and is used to train a model exhibiting better generalization towards novel classes.  ...  Acknowledgement: We gratefully acknowledge Brijesh Pillai and Partha Bhattacharya at Mercedes-Benz R&D India, Bangalore for providing the funding and infrastructure for this work.  ... 
arXiv:1909.07945v1 fatcat:5xuuyvqbpfe6neln7rgvhbfyke

Generalized Zero-Shot Learning for Action Recognition with Web-Scale Video Data [article]

Kun Liu, Wu Liu, Huadong Ma, Wenbing Huang, Xiongxiong Dong
2017 arXiv   pre-print
Then, we propose a method for action recognition by deploying generalized zero-shot learning, which transfers the knowledge of web video to detect the anomalous actions in surveillance videos.  ...  Moreover, it is very hard to collect real-word videos for certain particular actions such as steal and street fight due to legal restrictions and privacy protection.  ...  Generalized zero-shot learning. Zero-shot learning for action recognition in video has been widely studied in the computer version community.  ... 
arXiv:1710.07455v1 fatcat:datwl63c5jd2hiylkz7636lra4

Performance Improvement of Hand Gesture Recognition By using Sparse Coding With Kinect V2 Sensors

Here the important technique which is sparse coding representation acts as a major ROLE in achiveving One-short learning and real recognition of actions[1].  ...  to stand for 3D proceedings [2] .From this paper we obtain very good results in an domestic dataset captured by Kinect V2 sensors together with hand gesture proceedings and complex hand gesture actions  ...  But best of our knowledge any one of them fulfills the one-shot learning and high-level accuracy performance important for real life application Kinect V2 sensors [7] .  ... 
doi:10.35940/ijitee.k1976.0981119 fatcat:cycc37cop5fujdhchnnk6gvxra

Keep It Simple and Sparse: Real-Time Action Recognition [chapter]

Sean Ryan Fanello, Ilaria Gori, Giorgio Metta, Francesca Odone
2017 Gesture Recognition  
In this paper we show that sparse representation plays a fundamental role in achieving one-shot learning and real-time recognition of actions.  ...  We obtain very good results on three different data sets: a benchmark data set for one-shot action learning (the ChaLearn Gesture Data Set), an in-house data set acquired by a Kinect sensor including complex  ...  Although the presented algorithm leads to very good classification performance, it requires a computationally expensive offline learning phase that cannot be used in real-time for one-shot learning of  ... 
doi:10.1007/978-3-319-57021-1_10 fatcat:yxitbm4nsnhhjlef3azje4j2zi

Predicting Actions in Videos and Action-Based Segmentation Using Deep Learning

Fayaz A. Memon, Umair A. Khan, Asadullah Shaikh, Abdullah Alghamdi, Pardeep Kumar, Mesfer Alrizq
2021 IEEE Access  
ACKNOWLEDGMENT Authors would like to acknowledge the support of the Deputy for Research and Innovation-Ministry of Education, Kingdom of Saudi Arabia for this research through a grant (NU/IFC/INT/01/008  ...  Action recognition is useful in many applications including real-time surveillance [2] [3] , crowd behavior monitoring [4] [5] , bio-mechanical analysis of actions for sports and medicine [6]  ...  These events either entail real-time surveillance and response or require event-based archiving for later reference.  ... 
doi:10.1109/access.2021.3101175 fatcat:ekm7qofqnrfpljnlon34psdsfe

Semi-Supervised Few-Shot Atomic Action Recognition [article]

Xiaoyuan Ni, Sizhe Song, Yu-Wing Tai, Chi-Keung Tang
2020 arXiv   pre-print
To address the above issues, we focus on atomic actions and propose a novel model for semi-supervised few-shot atomic action recognition.  ...  Despite excellent progress has been made, the performance on action recognition still heavily relies on specific datasets, which are difficult to extend new action classes due to labor-intensive labeling  ...  Finally, the CTC and MSE loss enables our model for time-invariant few shot classification training. few-shot atomic action classification, that supports action recognition of long query videos under the  ... 
arXiv:2011.08410v1 fatcat:m4ezqx6pg5fs5kplm3fkviirg4

Scaling Human-Object Interaction Recognition in the Video through Zero-Shot Learning

Vali Ollah Maraghi, Karim Faez, Miguel Cazorla
2021 Computational Intelligence and Neuroscience  
We propose an approach for scaling human-object interaction recognition in video data through the zero-shot learning technique to solve this problem.  ...  Many successful works have been done on human-object interaction (HOI) recognition and achieved acceptable results in recent years.  ...  Some works focused on processing time that is appropriate for real-time object recognition tasks.  ... 
doi:10.1155/2021/9922697 fatcat:b6a73bphufcbzjssfyjocs4m4i

Exploring Relations in Untrimmed Videos for Self-Supervised Learning [article]

Dezhao Luo, Bo Fang, Yu Zhou, Yucan Zhou, Dayan Wu, Weiping Wang
2020 arXiv   pre-print
We validate our learned models with action recognition and video retrieval tasks with three kinds of 3D CNNs.  ...  Existing video self-supervised learning methods mainly rely on trimmed videos for model training. However, trimmed datasets are manually annotated from untrimmed videos.  ...  As shown in Table II , to clearly show the effect of relations for representation learning, we conduct ablation experiments on ERUV with various relations for action recognition.  ... 
arXiv:2008.02711v1 fatcat:zexhmmtazffyjepddxasb2dfoa

Football Players' Shooting Posture Norm Based on Deep Learning in Sports Event Video

Guangliang Huang, Zhuangxu Lan, Guo Huang, Le Sun
2021 Scientific Programming  
How to apply deep learning optimization to shooting gesture recognition is a very promising research direction.  ...  This article aims to study the football player's shooting posture specification based on deep learning in sports event videos.  ...  shot Outside instep Arch shot Forefoot shot Toe shot shot type Accuracy Table 2 : 2 Accuracy of various types of action recognition.  ... 
doi:10.1155/2021/1552096 fatcat:3vw5sxliarh4nl6oifzzwqbwji

TA2N: Two-Stage Action Alignment Network for Few-shot Action Recognition [article]

Shuyuan Li, Huabin Liu, Rui Qian, Yuxi Li, John See, Mengjuan Fei, Xiaoyuan Yu, Weiyao Lin
2021 arXiv   pre-print
Extensive experiments on benchmark datasets show the potential of the proposed method in achieving state-of-the-art performance for few-shot action recognition.  ...  Few-shot action recognition aims to recognize novel action classes (query) using just a few samples (support).  ...  The majority of current studies on few-shot action recognition follow the metric learning paradigm.  ... 
arXiv:2107.04782v2 fatcat:qqa4ymll2zhgvnw5qks7qz4tp4

Recognition of Badminton Shot Action Based on the Improved Hidden Markov Model

Chao Ma, Dayang Yu, Hao Feng, Fazlullah Khan
2021 Journal of Healthcare Engineering  
The experimental results show that the model designed can recognize ten standard strokes in real time.  ...  Among them, badminton has become one of the most popular sports because of the advantages of fewer restrictions on the field and ease of learning.  ...  Effect of Window Length on Recognition Performance. For badminton action recognition, fast response time is desired for real-time applications.  ... 
doi:10.1155/2021/7892902 pmid:34659693 pmcid:PMC8516566 fatcat:mgtt6n5isnar3b5yuvk2bmpyya

A Transfer Learning Method for Goal Recognition Exploiting Cross-Domain Spatial Features [article]

Thibault Duhamel, Mariane Maynard, Froduald Kabanza
2019 arXiv   pre-print
The trend, nowadays, is increasingly focusing on learning to infer intentions directly from data, using deep learning in particular.  ...  long sequences of actions to achieve their goals.  ...  Acknowledgements We thank Compute Canada for the computing resources they provided to support our research project.  ... 
arXiv:1911.10134v1 fatcat:62vkynje3zahxhzg5oebejmrsm

Deep Edge Computing for Videos

Jun-Hwa Kim, Namho Kim, Chee Sun Won
2021 IEEE Access  
So, they are suitable for the real-time applications.  ...  For example, in [24] , a two-path CNN model was proposed for video action recognition.  ... 
doi:10.1109/access.2021.3109904 fatcat:ctfwuywyfjh2rg74vxqkp2zf44
« Previous Showing results 1 — 15 out of 51,754 results