37,082 Hits in 5.9 sec


Valentin Radu, Maximilian Henne
2019 Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies  
The performance of a proposed network is first evaluated on a larger dataset, followed by transferring the pre-trained model to be fine-tuned on our five class activity recognition task.  ...  We show that transfer learning is also beneficial to identifying the best Convolutional Neural Network for vision based human activity recognition for our task.  ...  Vision2Sensor: Knowledge Transfer Across Sensing Modalities for Human Activity Recognition VALENTIN RADU, University of Edinburgh, UK MAXIMILIAN HENNE, University of Edinburgh, UK Mobile and wearable sensing  ... 
doi:10.1145/3351242 fatcat:nghrouxhfrbdvkwdrkdd4bdc2e

Transfer learning for activity recognition: a survey

Diane Cook, Kyle D. Feuz, Narayanan C. Krishnan
2013 Knowledge and Information Systems  
We characterize existing approaches to transfer-based activity recognition by sensor modality, by differences between source and target environments, by data availability, and by type of information that  ...  At the core of this capability is activity recognition, which is a challenging and well-researched problem.  ...  More work is needed to improve transfer across sensor modalities and to transfer knowledge across multiple differences.  ... 
doi:10.1007/s10115-013-0665-3 pmid:24039326 pmcid:PMC3768027 fatcat:eympkydr6vhrpdvbtgyimuvinq

ActiLabel: A Combinatorial Transfer Learning Framework for Activity Recognition [article]

Parastoo Alinia, Iman Mirzadeh, Hassan Ghasemzadeh
2020 arXiv   pre-print
However, an unprecedented increase in the diversity of sensor devices in the Internet-of-Things era has limited the adoption of activity recognition models for use across different domains.  ...  Sensor-based human activity recognition has become a critical component of many emerging applications ranging from behavioral medicine to gaming.  ...  More advanced models combine knowledge of transfer and deep learning . There have been studies that transfer different layers of deep neural networks across different domains.  ... 
arXiv:2003.07415v1 fatcat:6wifjqwlsnbuddgtc7chqz5lri

Transfer Learning beyond Text Classification [chapter]

Qiang Yang
2009 Lecture Notes in Computer Science  
Transfer Non-transfer Accuracy under 3m error distance 70% 65% 19 Outline 20  Transfer Learning in Sensor NetworksActivity Recognition  Heterogeneous Transfer Learning for Image  ...  of user moves Prior knowledge on the likelihood of where the user is Transfer No-transfer Accuracy under 3m error distance 85% 73% 17 Transferring Localization Models Across Space Transferring  ...  loc) will be a precondition of board-truck.  Conclusion:  The transferred knowledge help select a subset of candidate formulas, in case of learning from all the candidate formulas, which will make the  ... 
doi:10.1007/978-3-642-05224-8_3 fatcat:7tuzkm3rafdlroisihy6vcasqm

Cross-domain activity recognition via transfer learning

Derek Hao Hu, Vincent Wenchen Zheng, Qiang Yang
2011 Pervasive and Mobile Computing  
We show that after using this transfer learning approach, the performance of activity recognition in the new domain is increased several fold as compared to when no knowledge transfer is done.  ...  In this paper, we analyze the possibility of transferring knowledge from the available labeled data on a set of existing activities in one domain to help recognize the activities in another different but  ...  Acknowledgement We thank the support from Microsoft Research Asia and the support of project Hong Kong CERG/China-NSFC Grant N_HKUST624/09.  ... 
doi:10.1016/j.pmcj.2010.11.005 fatcat:p7jeh6zlsbgijeqwyhxcupz2de

Deep Transfer Learning for Cross-domain Activity Recognition [article]

Jindong Wang, Vincent W. Zheng, Yiqiang Chen, Meiyu Huang
2018 arXiv   pre-print
After this, we propose an effective Transfer Neural Network to perform knowledge transfer for Activity Recognition (TNNAR).  ...  activity knowledge transfer.  ...  Transfer Neural Network A er obtaining the selected source domains D s (K), we can perform knowledge transfer across the target and the source domains.  ... 
arXiv:1807.07963v2 fatcat:s2gmkcwkybfqtc5da32zkzgss4

Handling Variable-Dimensional Time Series with Graph Neural Networks [article]

Vibhor Gupta, Jyoti Narwariya, Pankaj Malhotra, Lovekesh Vig, Gautam Shroff
2020 arXiv   pre-print
We evaluate the proposed approach on publicly available activity recognition and equipment prognostics datasets, and show that the proposed approach allows for better generalization in comparison to a  ...  We propose a novel neural network architecture suitable for zero-shot transfer learning allowing robust inference for multivariate time series with previously unseen combination of available dimensions  ...  This approach considers knowledge transfer based on the similarity of installed sensors on body parts.  ... 
arXiv:2007.00411v5 fatcat:so2uzyw2jrf3vgg4ygvvxr5nj4

Semantics-aware Adaptive Knowledge Distillation for Sensor-to-Vision Action Recognition [article]

Yang Liu, Keze Wang, Guanbin Li, Liang Lin
2021 arXiv   pre-print
In this paper, we propose a novel framework, named Semantics-aware Adaptive Knowledge Distillation Networks (SAKDN), to enhance action recognition in vision-sensor modality (videos) by adaptively transferring  ...  Finally, to fully exploit and transfer the knowledge of multiple well-trained teacher networks to the student network, we propose a novel Graph-guided Semantically Discriminative Mapping loss, which utilizes  ...  Different from existing knowledge distillation methods that focus on the modality transfer task across vision-sensor based modalities, we move a further step towards knowledge transfer from wearable-sensor  ... 
arXiv:2009.00210v5 fatcat:pu6hlczumnbujmunvzuq7imbcq

Literature Review on Transfer Learning for Human Activity Recognition Using Mobile and Wearable Devices with Environmental Technology

Netzahualcoyotl Hernandez, Jens Lundström, Jesus Favela, Ian McChesney, Bert Arnrich
2020 SN Computer Science  
Thus, the importance of taking advantage of transfer learning to advance the field of activity recognition.  ...  Unsupervised learning across users is a typical application, and shallow techniques and active learning are areas of opportunity in transfer learning methodologies.  ...  possibility of taking advantage of the synergy of transfer knowledge across different technologies.  ... 
doi:10.1007/s42979-020-0070-4 fatcat:3opo6em35jfi5k5k3dwnrgxfse

Activity Recognition: Linking Low-level Sensors to High-level Intelligence

Qiang Yang
2009 International Joint Conference on Artificial Intelligence  
to be transformed into symbolic logical representations for AI planning, which in turn controls the robots or guides human users through a sensor network, thus completing a full life cycle of knowledge  ...  Building an activity recognition system requires a full range of interaction from statistical inference on lower level sensor data to symbolic AI at higher levels, where prediction results and acquired  ...  This semantic manifold acts as a bridge that propagates the common knowledge across different areas. [ Zheng et al., 2008a] considered transfer learning across sensor devices for a two-dimensional WiFi-based  ... 
dblp:conf/ijcai/Yang09 fatcat:fcbmme7wtbas5pha7g3ijo6orm

Wearable Computing

Daniel Roggen, Stéphane Magnenat, Markus Waibel, Gerhard Tröster
2011 IEEE robotics & automation magazine  
Acknowledgments We acknowledge the financial support of EU FP7 under the project OPPORTUNITY with grant number 225938 and RoboEarth with grant number 248942.  ...  Advances in merging concepts in ontologies [62] , [63] support the transfer of activity-recognition reasoning across different conceptual spaces.  ...  In the context of networked robots, activity recognition may help robots identify semantic relevance of their peer's activities, enhancing collaboration in heterogeneous robot teams.  ... 
doi:10.1109/mra.2011.940992 fatcat:pkfiwns5pbhm3aykxqh532nr34

A Survey of Vision-Based Transfer Learning in Human Activity Recognition

David Ada Adama, Ahmad Lotfi, Robert Ranson
2021 Electronics  
Human activity recognition (HAR) and transfer learning (TL) are two broad areas widely studied in computational intelligence (CI) and artificial intelligence (AI) applications.  ...  This paper presents a comprehensive survey of vision-based HAR using different methods with a focus on the incorporation of TL in HAR methods.  ...  TL, as an alternative to traditional learning methods, exists to aid in the transfer of knowledge across different but related situations of learning, so as to reuse knowledge and avoid having to train  ... 
doi:10.3390/electronics10192412 fatcat:offomcfmhvbdzasnowqlozm2ja

Three challenges in data mining

Qiang Yang
2010 Frontiers of Computer Science in China  
These challenges include: the transfer learning challenge, the social learning challenge and the mobile context mining challenge.  ...  I also believe that each of the three challenges discussed in this article will help move the science and engineering of data mining forward, and have a great impact on society.  ...  Activity recognition aims to understand a user's activities from a minimal set of available sensors at different levels of granularity.  ... 
doi:10.1007/s11704-010-0102-7 fatcat:o5sldxk2nbav3ldphvtqfverei

Instilling Social to Physical: Co-Regularized Heterogeneous Transfer Learning

Ying Wei, Yin Zhu, Cane Leung, Yangqiu Song, Qiang Yang
This rich set of socially shared activities motivates us to transfer knowledge from social media to address the sparsity issue of labelled physical sensor data.  ...  Ubiquitous computing tasks, such as human activity recognition (HAR), are enabling a wide spectrum of applications, ranging from healthcare to environment monitoring.  ...  Heterogeneous Transfer Learning (Yang et al. 2009 ) first proposed heterogeneous transfer learning which transfers knowledge across domains in different feature spaces.  ... 
doi:10.1609/aaai.v30i1.10172 fatcat:yorxb5tc7bdsfoz72h4q6tkha4

A Machine Vision Approach to Human Activity Recognition using Photoplethysmograph Sensor Data

Eoin Brophy, Jose Juan Dominguez, Zhengwei Wang, Tomas E. Ward
2018 2018 29th Irish Signals and Systems Conference (ISSC)  
Human activity recognition (HAR) is an active area of research concerned with the classification of human motion.  ...  Specifically, time-series images of photoplethysmography signals are used to retrain the penultimate layer of a pretrained convolutional neural network leveraging the concept of transfer learning.  ...  Of course, the addition of an inertial sensor will always produce more information and therefore more nuanced activity recognition.  ... 
doi:10.1109/issc.2018.8585372 fatcat:ohntcameujctnaqqsgkcz24odu
« Previous Showing results 1 — 15 out of 37,082 results