2021
DOI: 10.2339/politeknik.632070
|View full text |Cite
|
Sign up to set email alerts
|

Comparison of Methods for Determining Activity from Physical Movements

Abstract: The performances of the approaches that can recognize the physical movements of people are evaluated ❖ Accelerometer data of 6 different actions from 10 different people are obtained (from IMU) ❖ Accelerometer data ise divided into packets and their properties are extracted and classical approaches ❖ A new CNN-based approach to action detection is proposed ❖ Down, up, sitting, standing, walking and running actions were applied to YSA, SVM, k-NN and CNN

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

0
2
0
3

Year Published

2023
2023
2024
2024

Publication Types

Select...
6

Relationship

2
4

Authors

Journals

citations
Cited by 6 publications
(5 citation statements)
references
References 23 publications
0
2
0
3
Order By: Relevance
“…Since 2012, CNN architectures have dominated the field of machine learning, and most classical image processing and computer vision techniques have been redesigned with these architectures. The success of CNN has been experienced in various fields, particularly in image classification [20], object detection [21], scene classification [22], activity recognition from physical movements [23], Covid-19 detection from X-ray and CT images [24], brain MRI segmentation [25], brain tumor diagnosis [26], ECG arrhythmia classification [27], Covid-19 diagnosis from cough sound [28], Parkinson's disease detection from speech signals [29], etc., significantly improving the performance in these areas.…”
Section: Proposed Cnn Architecturementioning
confidence: 99%
“…Since 2012, CNN architectures have dominated the field of machine learning, and most classical image processing and computer vision techniques have been redesigned with these architectures. The success of CNN has been experienced in various fields, particularly in image classification [20], object detection [21], scene classification [22], activity recognition from physical movements [23], Covid-19 detection from X-ray and CT images [24], brain MRI segmentation [25], brain tumor diagnosis [26], ECG arrhythmia classification [27], Covid-19 diagnosis from cough sound [28], Parkinson's disease detection from speech signals [29], etc., significantly improving the performance in these areas.…”
Section: Proposed Cnn Architecturementioning
confidence: 99%
“…Çok kişili poz takibi insan hareketlerinin tanınması, video özetlemede önemli sahnelerin çıkartılması ve video içeriklerinin sınıflandırılmasında kullanılan önemli bir problemdir [1][2][3][4]. Poz takibi, insan eklem noktalarının yörüngelerinin doğru tahmini, insan eylemi tanıma, insan etkileşimini anlama, hareket yakalama ve animasyonlar için faydalıdır.…”
Section: Gi̇ri̇ş (Introduction)unclassified
“…Görüntü tabanlı veri setlerinde çoğunlukla derin öğrenme tabanlı yöntemler tercih edilirken, sayısal nitelikler içeren veri setlerinde ve işlem maliyetinin düşük olması istenen çalışmalarda Destek Vektör Makineleri (Support Vector Machine), k-En Yakın Komşu (k-nearest neighbors), karar ağacı vb. yöntemler tercih edilmektedir[1,[3][4][5][6][18][19][20].…”
unclassified