8,930 Hits in 3.5 sec

Human Activity Recognition with Deep Metric Learners

Kyle Martin, Anjana Wijekoon, Nirmalie Wiratunga
2019 International Conference on Case-Based Reasoning  
This is particularly important in the Human Activity Recognition (HAR) domain, where understanding similarity between cases supports aspects such as personalisation and open-ended HAR.  ...  Deep Metric Learners (DMLs) are a group of neural network architectures which learn to optimise case representations for similarity-based return by training upon multiple cases simultaneously to incorporate  ...  However, recent research has demonstrated that DMLs have great potential within the Human Activity Recognition (HAR) domain.  ... 
dblp:conf/iccbr/MartinWW19 fatcat:56mkhyhqifecpkypw2lmnwo7i4

Joint Learning of Temporal Models to Handle Imbalanced Data for Human Activity Recognition

Rebeen Ali Hamad, Longzhi Yang, Wai Lok Woo, Bo Wei
2020 Applied Sciences  
Moreover, imbalanced human activity datasets with less frequent activities create extra challenges for accurate activity recognition.  ...  However, it is challenging to deliver a sufficiently robust human activity recognition system from raw sensor data with noise in a smart environment setting.  ...  Several deep learning methods have been proposed for human activity recognition based on temporal data, which focuses on data processing.  ... 
doi:10.3390/app10155293 fatcat:b2eeu6pyy5e7vc4guc72mjch5y

Lifelong Ensemble Learning based on Multiple Representations for Few-Shot Object Recognition [article]

Hamidreza Kasaei, Songsong Xiong
2022 arXiv   pre-print
In particular, we form ensemble methods based on deep representations and handcrafted 3D shape descriptors.  ...  Service robots are integrating more and more into our daily lives to help us with various tasks.  ...  Active Object Recognition The performance of the deep 3D object recognition approach is heavily dependent on the quantity and quality of training data.  ... 
arXiv:2205.01982v3 fatcat:ydnphvz7jberrh5gntgslygxwy

Meta Multi-Task Learning for Speech Emotion Recognition

Ruichu Cai, Kaibin Guo, Boyan Xu, Xiaoyan Yang, Zhenjie Zhang
2020 Interspeech 2020  
Xia et al. use activation and valence information to help recognize categorical emotional labels based on the deep belief network with multi-task learning [11] .  ...  Most existing Speech Emotion Recognition (SER) approaches ignore the relationship between the categorical emotional labels and the dimensional labels in valence, activation or dominance space.  ...  Human emotions are complex and related to the way humans express emotions [9, 10] , such as valence (V, positive or negative), activation (A, calm or excited), and dominance (D, passive or aggressive)  ... 
doi:10.21437/interspeech.2020-2624 dblp:conf/interspeech/CaiGXYZ20 fatcat:wezru5oqzzgwfolsfehsu2eu2y

Multi-level Similarity Learning for Low-Shot Recognition [article]

Hongwei Xv, Xin Sun, Junyu Dong, Shu Zhang, Qiong Li
2019 arXiv   pre-print
According to this concept, we propose a multi-level similarity model (MLSM) to capture the deep encoded distance metric between the support and query samples.  ...  Low-shot learning indicates the ability to recognize unseen objects based on very limited labeled training samples, which simulates human visual intelligence.  ...  Compared with the image-level deep metric method RelationNets [26] , it can be noticed that the multi-level metric could be beneficial to learn a better similarity function with no additional training  ... 
arXiv:1912.06418v1 fatcat:vivmcpoq4fhqrdixwykq66jq6a

Deep Learning: The Impact on Future eLearning

Anandhavalli Muniasamy, Areej Alasiry
2020 International Journal of Emerging Technologies in Learning (iJET)  
It offers online learners of the future with intuitive algorithms and automated delivery of eLearning content through modern LMS platforms.  ...  trends of deep learning in eLearning, the relevant deep learning-based artificial intelligence tools and a platform enabling the developer and learners to quickly reuse resources are clearly summarized  ...  By using deep learning algorithm, artificial intelligence has a big breakthrough in many areas, such as face recognition [40] , image processing [3] , and speech recognition [2] .  ... 
doi:10.3991/ijet.v15i01.11435 fatcat:uj63cq5l7fchtbvipi7xyr4mfm

Exploring Inter-Observer Differences in First-Person Object Views Using Deep Learning Models

Chen Yu, Sven Bambach, Zehua Zhang, David J. Crandall
2017 2017 IEEE International Conference on Computer Vision Workshops (ICCVW)  
Recent advances in wearable camera technology have led many cognitive psychologists to study the development of the human visual system by recording the field of view of infants and toddlers.  ...  We consider a dataset of first-person videos from different people freely interacting with a set of toy objects, and train different objectrecognition models based on each subject's view.  ...  It used the FutureSystems Deep Learning facility, which is supported in part by IU and the NSF (RaPyDLI-1439007).  ... 
doi:10.1109/iccvw.2017.326 dblp:conf/iccvw/YuBZC17 fatcat:oxu7p2mlqfc5zcgh7ogrw63v5i

Sensor-based Human Activity Recognition using Deep Stacked Multilayered Perceptron Model

Furqan Rustam, Aijaz Ahmad Reshi, Imran Ashraf, Arif Mehmood, Saleem Ullah, Dost Muhammad Khan, Gyu Sang Choi
2020 IEEE Access  
CONCLUSION The study proposes a Deep Stacked Multilayered Perceptron (DS-MLP) model for human activity recognition tasks.  ...  Here dataset 1 is the 'Human Activity Recognition Using Smartphones Data Set' and dataset 2 is the 'Heterogeneity Human Activity Recognition Dataset'.  ... 
doi:10.1109/access.2020.3041822 fatcat:4ubdtevtxfbv5f5mfht3myjqmi

A Weighted Late Fusion Framework for Recognizing Human Activity from Wearable Sensors

Athina Tsanousa, Georgios Meditskos, Stefanos Vrochidis, Ioannis Kompatsiaris
2019 Zenodo  
This paper proposes a novel framework for combining accelerometers and gyroscopes at decision level, in order to recognize human activity.  ...  recognition rate.  ...  In summary, a human activity recognition framework begins with the collection or extraction of raw data, like sensor signals or video images.  ... 
doi:10.5281/zenodo.3507004 fatcat:havv2ws5wfadndleakio75zswq

Deep Learning Analysis of Mobile Physiological, Environmental and Location Sensor Data for Emotion Detection

Eiman Kanjo, Eman M.G. Younis, Chee Siang Ang
2018 Information Fusion  
In recent years, deep learning has been increasingly used in the field of human activity recognition [17, 21] . While progress has been made, human activity recognition remains a challenging task.  ...  Since deep learning is capable of high-level abstraction of data, it can be used to develop self-configurable frameworks for human activity as well as emotion recognition.  ...  Supplementary material Supplementary material associated with this article can be found, in the online version, at doi: 10.1016/j.inffus.2018.09.001.  ... 
doi:10.1016/j.inffus.2018.09.001 fatcat:di56h2kpcjcdhondjlyghm2gai

Recognition physical activities with optimal number of wearable sensors using data mining algorithms and deep belief network

Ali H. Al-Fatlawi, Hayder K. Fatlawi, Sai Ho Ling
2017 2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)  
Furthermore, the deep belief network (DBN) has been investigated to recognise 33 physical activities effectively.  ...  Daily physical activities monitoring is benefiting the health care field in several ways, in particular with the development of the wearable sensors.  ...  Apart from its benefits in health applications, recognition of the human physical activities by the wearable sensors can serve various kinds of sectors.  ... 
doi:10.1109/embc.2017.8037456 pmid:29060497 dblp:conf/embc/Al-FatlawiFL17 fatcat:npi4esst6jdglkvbcp7dokpvfa

Speech Processing for Language Learning: A Practical Approach to Computer-Assisted Pronunciation Teaching

Natalia Bogach, Elena Boitsova, Sergey Chernonog, Anton Lamtev, Maria Lesnichaya, Iurii Lezhenin, Andrey Novopashenny, Roman Svechnikov, Daria Tsikach, Konstantin Vasiliev, Evgeny Pyshkin, John Blake
2021 Electronics  
The CRQA metrics combined with those of DTW were shown to add to the accuracy of learner performance estimation.  ...  We also examine the scope of automatic speech recognition applicability within the CAPT system workflow and evaluate the Levenstein distance between the transcription made by human experts and that obtained  ...  As such, using deep neural networks for recognition of learner speech might provide a sufficiently accurate result to be used to echo learner's speech input and can be displayed along with the other visuals  ... 
doi:10.3390/electronics10030235 fatcat:io6bpmxglbeo5jp4bbz6depy5e

Sketch Recognition with Deep Visual-Sequential Fusion Model

Jun-Yan He, Xiao Wu, Yu-Gang Jiang, Bo Zhao, Qiang Peng
2017 Proceedings of the 2017 ACM on Multimedia Conference - MM '17  
In this paper, a deep end-to-end network for sketch recognition, named Deep Visual-Sequential Fusion model (DVSF) is proposed to model the visual and sequential pa erns of the strokes.  ...  To capture the intermediate states of sketches, a three-way representation learner is rst utilized to extract the visual features. ese deep features are simultaneously fed into the visual and sequential  ...  Evaluation Metric.  ... 
doi:10.1145/3123266.3123321 dblp:conf/mm/HeWJZP17 fatcat:pbbvhkyqwze67j2y3cyluyleqm

Deep Multiple Metric Learning for Time Series Classification

Zhi Chen, Yongguo Liu, Jiajing Zhu, Yun Zhang, Qiaoqin Li, Rongjiang Jin, Xia He
2021 IEEE Access  
INDEX TERMS Adversarial training, deep learning, metric learning, time series classification.  ...  learners.  ...  To show the effectiveness of DMML on heterogeneous dataset, we conduct experiments on Heterogeneity Human Activity Recognition (HHAR) dataset [53] .  ... 
doi:10.1109/access.2021.3053703 fatcat:erhjde4xmfbdjokuqmhrkmo5by

Impact of Deep Learning on Transfer Learning : A Review

M. J. Barwary, A. M. Abdulazeez
2021 Zenodo  
Transfer learning and deep learning approaches have been utilised in several real-world applications and hierarchical systems for pattern recognition and classification tasks.  ...  competent with more easily attained data from diverse fields.  ...  Eaton's paper (Eaton & Lane, 2008) proposes to create a target learner based on a transferability metric from several linked source domains.  ... 
doi:10.5281/zenodo.4559668 fatcat:2sqju4jirbd5jae3gccd3kkm6y
« Previous Showing results 1 — 15 out of 8,930 results