66,067 Hits in 6.2 sec

Complex Deep Neural Networks from Large Scale Virtual IMU Data for Effective Human Activity Recognition Using Wearables

Hyeokhyen Kwon, Gregory D. Abowd, Thomas Plötz
2021 Sensors  
Supervised training of human activity recognition (HAR) systems based on body-worn inertial measurement units (IMUs) is often constrained by the typically rather small amounts of labeled sample data.  ...  Our models contain components that are dedicated to capture the essentials of IMU data as they are of relevance for activity recognition, which increased the number of trainable parameters by a factor  ...  Deriving human activity recognition systems from large scale virtual IMU data generated through IMUTube [14] .  ... 
doi:10.3390/s21248337 pmid:34960431 pmcid:PMC8707382 fatcat:ex27jvnfxnbq3g7b2wfdqmmy2a

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  
and interaction analysis X [32] Human Activity Recognition in Smart Homes X X Proposed Video-based Human Activity Recognition with deep learning A.  ...  The spatial pyramid module extracts multi-scale appearance features of video frames.  ... 
doi:10.1109/access.2021.3110610 fatcat:ussooxm7azfljpb5prsm7creaa

Directional Temporal Modeling for Action Recognition [article]

Xinyu Li, Bing Shuai, Joseph Tighe
2020 arXiv   pre-print
Our CIDC network can be attached to any activity recognition backbone network.  ...  Many current activity recognition models use 3D convolutional neural networks (e.g. I3D, I3D-NL) to generate local spatial-temporal features.  ...  The details of this multi-scale aggregation operator is illustrated in Figure 3 .  ... 
arXiv:2007.11040v1 fatcat:lkfdormidjfz7bqb3basuddmti

STAC: Spatial-Temporal Attention on Compensation Information for Activity Recognition in FPV

Yue Zhang, Shengli Sun, Linjian Lei, Huikai Liu, Hui Xie
2021 Sensors  
It generates a multi-scale set of regions, including multi-size objects, leading to superior performance. We compensate for the optical flow to eliminate the camera noise in motion.  ...  Egocentric activity recognition in first-person video (FPV) requires fine-grained matching of the camera wearer's action and the objects being operated.  ...  Acknowledgments: The authors would like to acknowledge the Georgia Institute of Technology for making their Egocentric Activity Datasets available.  ... 
doi:10.3390/s21041106 pmid:33562612 pmcid:PMC7914484 fatcat:ktcxuueywzb43fan7tra3cdi6i

MoFAP: A Multi-level Representation for Action Recognition

Limin Wang, Yu Qiao, Xiaoou Tang
2015 International Journal of Computer Vision  
This paper proposes a multi-level video representation by stacking the activations of motion features, atoms, and phrases (MoFAP).  ...  Motion atom is defined as an atomic part of action, and captures the motion information of video in a short temporal scale.  ...  We call these motion atoms discovered from different temporal scales as multi-scale motion atoms.  ... 
doi:10.1007/s11263-015-0859-0 fatcat:zxjizsbntzfm5kriz42qgxsi74

Author Index

2010 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition  
On Growth and Formlets: Sparse Multi-Scale Coding of Planar Shape Oliva, Aude SUN Database: Large Scale Scene Recognition from Abbey to Zoo Oliveira, Francisco Workshop: Using a Vision Based Tracking  ...  Gait Recognition based on Local Motion Feature Selection Multi-View Structure Computation without Explicitly Estimating Motion Li, Hongsheng Object Matching with a Locally Affine-Invariant Constraint  ... 
doi:10.1109/cvpr.2010.5539913 fatcat:y6m5knstrzfyfin6jzusc42p54

A Modified Residual Network Based on Multi-scale Segmentation for Aerobics Motion Image Recognition

Xingxing Dai Xingxing Dai
2022 Diànnǎo xuékān  
In this paper, we propose a modified Residual network based on multi-scale segmentation for aerobics motion image recognition.  ...  Then, combining with the multi-scale segmentation method, an image recognition residual network is obtained.  ...  Proposed Motion Image Recognition Method The multi-scale segmentation method of HS-ResNet enables different groups of feature information to enjoy different scales of receptive fields.  ... 
doi:10.53106/199115992022023301006 fatcat:5k53fryjargujfyhewpgso35va

Multi-Modal Emotion Recognition Fusing Video and Audio

Chao Xu, Pufeng Du, Zhiyong Feng, Zhaopeng Meng, Tianyi Cao, Caichao Dong
2013 Applied Mathematics & Information Sciences  
Experiments indicate that multi-modal fusion emotion recognition algorithm which is presented in this paper has relatively high recognition accuracy.  ...  We construct a framework for multi-modal fusion emotion recognition. Facial expression features and speech features are respectively extracted from image sequences and speech signals.  ...  As a result, emotion recognition is inherently an issue of multi-modal analysis.  ... 
doi:10.12785/amis/070205 fatcat:ardm56lkdvd6hm5e45mbytnhfy

Composite Feature Vector Assisted Human Action Recognition through Supervised Learning

2020 International journal of recent technology and engineering  
However, the existing approaches have many limitations like low recognition accuracy and non-robustness. Hence, this paper focused to develop a novel and robust Human Action Recognition framework.  ...  Experimental results reveal that the proposed framework achieves better performance compared to state-of-art recognition methods.  ...  [28] combined Gabor Transform with Ridgelet Transform to perform human activity recognition.  ... 
doi:10.35940/ijrte.f7337.038620 fatcat:j3jixe5ngvckrmba44243iygsy

Analysis of Multi-Scale Fractal Dimension to Classify Human Motion [article]

Núbia Rosa da Silva, Odemir Martinez Bruno
2012 arXiv   pre-print
The objective of this study is to investigate the use of 3D multi-scale fractal dimension to recognize motion patterns in videos.  ...  In recent years there has been considerable interest in human action recognition. Several approaches have been developed in order to enhance the automatic video analysis.  ...  Multi-scale fractal dimension curve is defined through the derivative of log V (r) × log r curve.  ... 
arXiv:1207.1649v1 fatcat:nj4zsjosffciviabdj2c5dtlva

Vector field analysis for multi-object behavior modeling

Nandita M. Nayak, Yingying Zhu, Amit K. Roy-Chowdhury
2013 Image and Vision Computing  
a single optical flow based motion analysis framework.  ...  Nayak, et al., Vector field analysis for multi-object behavior modeling, Image Vis. Comput. (2012), http:// dx.  ...  They could have multiple events occurring simultaneously at arbitrary viewpoints and varying scales. The analysis of such videos can be termed as complex activity recognition.  ... 
doi:10.1016/j.imavis.2012.08.011 fatcat:abljsz4uzncozber54ch2adcum

Information, communication and computing technologies as enablers of advancements in modern information society

Anton Kos, Yunchuan Sun, Rongfang Bie
2021 Personal and Ubiquitous Computing  
Such examples include new smart devices and various sensor systems used in motor learning and performance improvement in sports, as well as motion/condition recognition in ubiquitous healthcare.  ...  These functionalities are demonstrated through a case study from a real credit evaluation company.  ...  The design of multi-scale structure makes the analysis of surveillance content natural and fluent with the annotation of video content.  ... 
doi:10.1007/s00779-021-01557-w fatcat:kqatzspagrhadezg2jpcqz4z7e

Human Pose Estimation and Activity Recognition From Multi-View Videos: Comparative Explorations of Recent Developments

Michael B. Holte, Cuong Tran, Mohan M. Trivedi, Thomas B. Moeslund
2012 IEEE Journal on Selected Topics in Signal Processing  
This paper presents a review and comparative study of recent multi-view approaches for human 3D pose estimation and activity recognition.  ...  We report a comparison of the most promising methods for multi-view human action recognition using two publicly available datasets: the INRIA Xmas Motion Acquisition Sequences (IXMAS) Multi-View Human  ...  Instead of manually observing the data for analysis, such studies can utilize the recent advances in pose estimation and activity analysis to automate the process and enable analysis in a larger scale.  ... 
doi:10.1109/jstsp.2012.2196975 fatcat:2sqvghpigzczzlrrl3iziyg63a

Human activity recognition based on the blob features

Jie Yang, Jian Cheng, Hanqing Lu
2009 2009 IEEE International Conference on Multimedia and Expo  
In this paper, we present a novel approach for human activities recognition in the video.  ...  Firstly, we establish a statistical background model and extract foreground object through background subtraction in the video stream.  ...  INTRODUCTION Automatic human activities recognition in video streams is gaining more and more attention in the video analysis research community due to many video applications such as video content analysis  ... 
doi:10.1109/icme.2009.5202508 dblp:conf/icmcs/YangCL09 fatcat:5fdoz7wtpvdwxobklwtosav2wm

Multimodal Multi-Stream Deep Learning for Egocentric Activity Recognition

Sibo Song, Vijay Chandrasekhar, Bappaditya Mandal, Liyuan Li, Joo-Hwee Lim, Giduthuri Sateesh Babu, Phyo Phyo San, Ngai-Man Cheung
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
In this paper, we propose a multimodal multi-stream deep learning framework to tackle the egocentric activity recognition problem, using both the video and sensor data.  ...  Experimental results using a multimodal egocentric dataset show that our proposed method can achieve very encouraging performance, despite the constraint that the scale of the existing egocentric datasets  ...  Training deep ConvNets is more challenging for egocentric activity recognition as activity is more complex (considerable camera motion in addition to the object motion in the scene).  ... 
doi:10.1109/cvprw.2016.54 dblp:conf/cvpr/SongCMLLBSC16 fatcat:myz44dahlbavvio5syy4nlopby
« Previous Showing results 1 — 15 out of 66,067 results