2021 American Control Conference (ACC) 2021
DOI: 10.23919/acc50511.2021.9483252
|View full text |Cite
|
Sign up to set email alerts
|

Step Length Estimation Using Inertial Measurements Units

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
3
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
3
2
1

Relationship

0
6

Authors

Journals

citations
Cited by 7 publications
(3 citation statements)
references
References 13 publications
0
3
0
Order By: Relevance
“…To compare our LSTM algorithm to other standard classifiers commonly used to make predictions on large datasets, we have created receiver operating characteristic (ROC) curves for 6 activities of standing, walking, sit-stand transitions, turning, bending and near-falls. Figure 4 shows the ROC curves for four binary classifiers of Logistic Regression (LOG), Support Vector Machines (SVM), Decision Tree (DT), and our Long-Short-Term-Memory cells (LSTM) for each activity [15]. The area under the curve for each plot is summarized in Table 3.…”
Section: Resultsmentioning
confidence: 99%
See 1 more Smart Citation
“…To compare our LSTM algorithm to other standard classifiers commonly used to make predictions on large datasets, we have created receiver operating characteristic (ROC) curves for 6 activities of standing, walking, sit-stand transitions, turning, bending and near-falls. Figure 4 shows the ROC curves for four binary classifiers of Logistic Regression (LOG), Support Vector Machines (SVM), Decision Tree (DT), and our Long-Short-Term-Memory cells (LSTM) for each activity [15]. The area under the curve for each plot is summarized in Table 3.…”
Section: Resultsmentioning
confidence: 99%
“…In order to properly identify home activities, we first defined a variety of activities using the IMU data (Supplemental Table 1). The algorithm used to identify each home activity is a deep learning-based activity recognition architecture using a convolutional neural network long short-term memory network (CNN-LSTM) which we have previously detailed [15]. We also used three other commonly used classifiers (logistic regression, support vector machine, decision tree) to compare their performance to our CNN-LSTM.…”
Section: Methodsmentioning
confidence: 99%
“…Although step length and step time were initially included in our study for model training, they were not used to train the model as they were not selected by the feature-ranking technique. This inconsistency could possibly be derived from the inclusion of joint angles in our study, as step length could be accurately calculated based on the joint angles of the lower limbs, such as the hip and knee [ 33 ]. Rather than utilizing step length in our model, all lower-limb joint angles were incorporated (left and right hip, left and right knee).…”
Section: Discussionmentioning
confidence: 99%