Filters








19,938 Hits in 6.2 sec

Explaining Motion Relevance for Activity Recognition in Video Deep Learning Models [article]

Liam Hiley and Alun Preece and Yulia Hicks and Supriyo Chakraborty and Prudhvi Gurram and Richard Tomsett
2020 arXiv   pre-print
A small subset of explainability techniques developed initially for image recognition models has recently been applied for interpretability of 3D Convolutional Neural Network models in activity recognition  ...  In fact, it has been shown that these models do not appropriately factor motion information into their decision.  ...  Governments are authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation hereon.  ... 
arXiv:2003.14285v1 fatcat:24b2igwemnfche3pjsrq2rat3e

Explainable Deep Learning for Video Recognition Tasks: A Framework Recommendations [article]

Liam Hiley, Alun Preece, Yulia Hicks
2019 arXiv   pre-print
The current work in explainability for video models is still overshadowed by Image techniques, while Video Deep Learning itself is quickly gaining on methods for still images.  ...  This paper seeks to highlight the need for explainability methods designed with video deep learning models, and by association spatio-temporal input in mind, by first illustrating the cutting edge for  ...  Framework for Future Research In this paper we have addressed some gaps in the current works for explainable activity recognition with video deep learning models.  ... 
arXiv:1909.05667v1 fatcat:jnjri66khfgedn4hyxtqaw55g4

Attention is All We Need: Nailing Down Object-centric Attention for Egocentric Activity Recognition [article]

Swathikiran Sudhakaran, Oswald Lanz
2018 arXiv   pre-print
In this paper we propose an end-to-end trainable deep neural network model for egocentric activity recognition.  ...  We visually analyze attention maps generated by the network, revealing that the network successfully identifies the relevant objects present in the video frames which may explain the strong recognition  ...  Several deep learning based methods have been proposed for action recognition from third person videos in recent years.  ... 
arXiv:1807.11794v1 fatcat:sqentueowrba5oesvn2wgxxsky

Video Activity Recognition: State-of-the-Art

Itsaso Rodríguez-Moreno, José María Martínez-Otzeta, Basilio Sierra, Igor Rodriguez, Ekaitz Jauregi
2019 Sensors  
The aim of this paper is to survey the state-of-the-art techniques for video activity recognition while at the same time mentioning other techniques used for the same task that the research community has  ...  Video activity recognition, although being an emerging task, has been the subject of important research efforts due to the importance of its everyday applications.  ...  Conclusions In this paper, different methods for video activity recognition have been presented. Several models have been explained showing the development of recent years.  ... 
doi:10.3390/s19143160 fatcat:ore5paemmbfxzjimsvnnezgwwi

A Review on state-of-the-art Violence Detection Techniques

Muhammad Ramzan, Adnan Abid, Hikmat Ullah Khan, Shahid Mahmood Awan, Amina Ismail, Muzamil Ahmed, Mahwish Ilyas, Ahsan Mahmood
2019 IEEE Access  
The relevant literature presents different techniques for detection of such activities from the video proposed in the recent years.  ...  Moreover, datasets and video features that used in the techniques, which play a vital role in recognition process are also discussed.  ...  Inspired by the performance of deep models for the recognition of human action, an innovative method for the detection of human violent behavior by combining the trajectory and deep CNN is proposed [44  ... 
doi:10.1109/access.2019.2932114 fatcat:v5bkozfisneprkvikxpi4scnkq

Convolutional Long Short-Term Memory Networks for Recognizing First Person Interactions [article]

Swathikiran Sudhakaran, Oswald Lanz
2017 arXiv   pre-print
In this paper, we present a novel deep learning based approach for addressing the problem of interaction recognition from a first person perspective.  ...  by more than 20% in recognition accuracy.  ...  Acknowledgments We gratefully acknowledge the support of NVIDIA Corporation with the donation of GPU used for this research.  ... 
arXiv:1709.06495v1 fatcat:ikslmyzh2jednpwprc3ra6x5wi

Convolutional Long Short-Term Memory Networks for Recognizing First Person Interactions

Swathikiran Sudhakaran, Oswald Lanz
2017 2017 IEEE International Conference on Computer Vision Workshops (ICCVW)  
In this paper, we present a novel deep learning based approach for addressing the problem of interaction recognition from a first person perspective.  ...  by more than 20% in recognition accuracy.  ...  Acknowledgments We gratefully acknowledge the support of NVIDIA Corporation with the donation of GPU used for this research.  ... 
doi:10.1109/iccvw.2017.276 dblp:conf/iccvw/SudhakaranL17 fatcat:dmkd6noojjfvvekouzehqtuecy

Analysis of Deep Neural Networks For Human Activity Recognition in Videos – A Systematic Literature Review

Hadiqa Aman Ullah, Sukumar Letchmunan, M. Sultan Zia, Umair Muneer Butt, Fadratul Hafinaz Hassan
2021 IEEE Access  
-What deep learning techniques currently exist since 2015 to model Spatio-temporal information to support activity recognition in videos?  ...  [28] reviewed the activity recognition trends with deep learning models. This paper examined various models based on two-stream networks, C3D and RNN, used for activity recognition.  ...  She has previously served as a lecturer in the field of Computer Science and IT and has one journal publication before. Her research interests are Data Science, Machine learning, and Computer vision.  ... 
doi:10.1109/access.2021.3110610 fatcat:ussooxm7azfljpb5prsm7creaa

LSTA: Long Short-Term Attention for Egocentric Action Recognition

Swathikiran Sudhakaran, Sergio Escalera, Oswald Lanz
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Egocentric activity recognition is one of the most challenging tasks in video analysis. It requires a fine-grained discrimination of small objects and their manipulation.  ...  In this paper we propose LSTA as a mechanism to focus on features from relevant spatial parts while attention is being tracked smoothly across the video sequence.  ...  We gratefully acknowledge the support of NVIDIA Corporation with the donation of GPUs used for this research.  ... 
doi:10.1109/cvpr.2019.01019 dblp:conf/cvpr/SudhakaranEL19 fatcat:numtqwnpdjgijhao3s4e7optii

A Comprehensive Review of Group Activity Recognition in Videos

Li-Fang Wu, Qi Wang, Meng Jian, Yu Qiao, Bo-Xuan Zhao
2021 International Journal of Automation and Computing  
Second, we survey the group activity recognition methods, including those based on handcrafted features and those based on deep learning networks.  ...  In this paper, we give a comprehensive overview of the advances in group activity recognition in videos during the past 20 years.  ...  This section reviews deep learning based methods for group activity recognition.  ... 
doi:10.1007/s11633-020-1258-8 fatcat:ycka4thcy5a6vghpenpthtrndi

Explainable Video Action Reasoning via Prior Knowledge and State Transitions [article]

Tao Zhuo, Zhiyong Cheng, Peng Zhang, Yongkang Wong, Mohan Kankanhalli
2019 arXiv   pre-print
Our method takes advantage of both classical reasoning and modern deep learning approaches.  ...  Specifically, prior knowledge is defined as the information of a target video domain, including a set of objects, attributes and relationships in the target video domain, as well as relevant actions defined  ...  Recently, with the success of deep learning in a variety of computer vision tasks, great progress has been achieved in video action recognition with various deep neural networks [3, 9, 41, 44, 48] .  ... 
arXiv:1908.10700v1 fatcat:itbcavk37fgkfmnhn5syzze6iy

Video Representation via Fusion of Static and Motion Features Applied to Human Activity Recognition

2019 KSII Transactions on Internet and Information Systems  
In human activity recognition system both static and motion information play crucial role for efficient and competitive results.  ...  Second, trajectories are extracted by tracking key points and only those trajectories have been selected which are located in central region of the original video frame in order to to reduce irrelevant  ...  deep learning-based models for action recognition.  ... 
doi:10.3837/tiis.2019.07.015 fatcat:evwphdggdnbivek3axqxmwumn4

A Novel Recurrent Neural Networks Architecture for Behavior Analysis

Neziha Jaouedi, Noureddine Boujnah, Mohamed Bouhlel
2021 ˜The œinternational Arab journal of information technology  
In this paper, we propose new model for human behavior analysis which combine transfer learning model and Recurrent Neural Network (RNN).  ...  In fact, in the increase of crimes, everyone needs video surveillance to keep their belongings safe and to automatically detect events by collecting important information for the assistance of security  ...  Our proposed model for human action recognition relies on the extraction of relevant features from Inception v3.  ... 
doi:10.34028/iajit/18/2/1 fatcat:tm4mche26zezbljgr4dwcqtr3i

NTU RGB+D 120: A Large-Scale Benchmark for 3D Human Activity Understanding

Jun Liu, Amir Shahroudy, Mauricio Lisboa Perez, Gang Wang, Ling-Yu Duan, Alex Kot Chichung
2019 IEEE Transactions on Pattern Analysis and Machine Intelligence  
We evaluate the performance of a series of existing 3D activity analysis methods on this dataset, and show the advantage of applying deep learning methods for 3D-based human action recognition.  ...  In this work, we introduce a large-scale dataset for RGB+D human action recognition, which is collected from 106 distinct subjects and contains more than 114 thousand video samples and 8 million frames  ...  Recently, deep learning based-approaches have been proposed for 3D human activity analysis [9] , [59] , [60] .  ... 
doi:10.1109/tpami.2019.2916873 pmid:31095476 fatcat:xlwrbwwucvdjpofkt5lwdyqy4m

Automatic Unusual Activities Recognition Using Deep Learning in Academia

Muhammad Ramzan, Adnan Abid, Shahid Mahmood Awan
2022 Computers Materials & Continua  
The proposed model has been named Automatic Unusual Activity Recognition (AUAR), which employs motion-based frame extraction approaches to extract key-frames and then applies advanced deep learning Convolutional  ...  To the best of our knowledge, this is the first research work in this area that develops a dataset for unusual activities in the examination and proposes a deep learning model to detect those unusual activities  ...  The proposed system extracts key-frames based on human motion from a video sequence/stream; deep learning model 2D and 3D CNN used for classification task to detect suspicious activities.  ... 
doi:10.32604/cmc.2022.017522 fatcat:cqtmvtdj6fcr7lm33fzomjhswm
« Previous Showing results 1 — 15 out of 19,938 results