A Platform and Methodology Enabling Real-Time Motion Pattern Recognition on Low-Power Smart Devices

Sarbishei, O.

doi:10.1109/wf-iot.2019.8767219

Cited by 8 publications

(20 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: Resultscontrasting

confidence: 63%

“…Figure 1 a shows that the Feedforward Neural Network has a low F1 score (0.36) compared to other classifiers (above 0.5), which contradicts the results reported in [ 18 ] where the Feedforward Neural Network achieves more than 95% accuracy in a context of offline training. The main difference between [ 18 ] and our study lie in the definition of the training set. In [ 18 ], the training set includes examples from every subject, while we only use a single one, to ensure an objective comparison with the other stream classifiers that do not require offline training (except for hyperparameter tuning, done on the first subject of the Banos et al dataset).…”

Section: Resultscontrasting

confidence: 63%

“…Finally, Neural Network classifiers have become popular by reaching or even exceeding human performance in many fields such as image recognition or game playing. We use a Feedforward Neural Network with one hidden layer, as described in [ 18 ] for the recognition of fitness activities on a low-power platform.…”

Section: Methodsmentioning

confidence: 99%

“…In this study, we used histogram features from [ 18 ] instead of the ones presented in Section 3.1 because the network performed poorly with these features. The histogram features produce 20 bins per axis.…”

Section: Methodsmentioning

confidence: 99%

“…In a data stream (online) setting, the work in [ 18 ] presents a wearable system capable of running pre-trained classifiers on the chip with high classification accuracy. This shows the superiority of the proposed Feedforward Neural Network over k -NN.…”

Section: Related Workmentioning

confidence: 99%

See 4 more Smart Citations

A Benchmark of Data Stream Classification for Human Activity Recognition on Connected Objects

Khannouz

Glatard

2020

Sensors

View full text Add to dashboard Cite

This paper evaluates data stream classifiers from the perspective of connected devices, focusing on the use case of Human Activity Recognition. We measure both the classification performance and resource consumption (runtime, memory, and power) of five usual stream classification algorithms, implemented in a consistent library, and applied to two real human activity datasets and three synthetic datasets. Regarding classification performance, the results show the overall superiority of the Hoeffding Tree, the Mondrian forest, and the Naïve Bayes classifiers over the Feedforward Neural Network and the Micro Cluster Nearest Neighbor classifiers on four datasets out of six, including the real ones. In addition, the Hoeffding Tree and—to some extent—the Micro Cluster Nearest Neighbor, are the only classifiers that can recover from a concept drift. Overall, the three leading classifiers still perform substantially worse than an offline classifier on the real datasets. Regarding resource consumption, the Hoeffding Tree and the Mondrian forest are the most memory intensive and have the longest runtime; however, no difference in power consumption is found between classifiers. We conclude that stream learning for Human Activity Recognition on connected objects is challenged by two factors which could lead to interesting future work: a high memory consumption and low F1 scores overall.

show abstract

Section: Resultscontrasting

confidence: 63%

Section: Resultscontrasting

confidence: 63%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 3 more Smart Citations

A Benchmark of Data Stream Classification for Human Activity Recognition on Connected Objects

Khannouz

Glatard

2020

Sensors

View full text Add to dashboard Cite

show abstract

Application of Machine Learning-Based Pattern Recognition in IoT Devices: Review

Menter

Tee

Dave

2021

Algorithms for Intelligent Systems

View full text Add to dashboard Cite

The Internet of Things (IoT) is a rapidly advancing area of technology that has quickly become more widespread in recent years. With greater numbers of everyday objects being connected to the internet, many different innovations have been presented to make our everyday lives more straightforward. Pattern recognition is extremely prevalent in IoT devices because of the many applications and benefits that can come from it. A multitude of studies have been conducted with the intention of improving speed and accuracy, decreasing complexity, and reducing the overall required processing power of pattern recognition algorithms in IoT devices. After reviewing the applications of different machine learning algorithms, results vary from case to case but a general conclusion can be drawn that the optimal machine learning based pattern recognition algorithms to be used with IoT devices are K-Nearest Neighbor, Random Forest, and Support Vector Machine.

show abstract

An Analysis on Daily and Sports Activities’ Recognition Using Data Analytics and Artificial Intelligence Model

Maheswari

Radha

2023

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

A Platform and Methodology Enabling Real-Time Motion Pattern Recognition on Low-Power Smart Devices

Cited by 8 publications

References 8 publications

A Benchmark of Data Stream Classification for Human Activity Recognition on Connected Objects

A Benchmark of Data Stream Classification for Human Activity Recognition on Connected Objects

Application of Machine Learning-Based Pattern Recognition in IoT Devices: Review

An Analysis on Daily and Sports Activities’ Recognition Using Data Analytics and Artificial Intelligence Model

Contact Info

Product

Resources

About