Ensemble kNN classifiers for human gait recognition based on ground reaction forces

Derlatka, M.; Bogdan, Mariusz

doi:10.1109/hsi.2015.7170648

Cited by 28 publications

(11 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This may have reduced the variability of the data and in turn simplified the re-ID task (Table IV) [64]. The benchmark for single session re-ID was 97.38 % accuracy on 200 people, an additional 1.36 % compared to our best result on 118 people [65]. Still, this comparison is vague because the authors did not provide a detailed description of the experimental protocol or dataset characteristics.…”

Section: Discussionmentioning

confidence: 99%

The Most Discriminant Components of Force Platform Data for Gait Based Person Re-identification

Duncanson¹,

Thwaites²,

Booth³

et al. 2021

Preprint

View full text Add to dashboard Cite

Walking gait data measured using force platforms is a promising means for person re-identification in authentication and surveillance scenarios. We aimed to determine the most discriminant components of force platform data using a two-stream Convolutional Recurrent Neural Network (KineticNet). Each network in the two-stream architecture extracts features pertaining to a single stance phase and then these features are fused to represent the entire gait cycle. Over two sessions, ground reaction forces (Fx, Fy, Fz), moments (Mx, My, Mz), and center of pressure coordinates (Cx, Cy) were acquired from 118 participants as they walked our laboratory five times at preferred speed. For each participant and each session, up to three samples were reserved for network training, leaving one sample for network validation and one sample for network testing. KineticNet’s performance was evaluated using both individual component and multi-component inputs before ablation studies were conducted on its architecture. Fz was the most discriminant individual component, and re-identification using Fz, Fy, and Cy together was the most accurate overall at 96.02%. These results warrant further investigation into the utility of force platforms as an accessory or alternative to video cameras for gait based person re-identification.

show abstract

Section: Discussionmentioning

confidence: 99%

The Most Discriminant Components of Force Platform Data for Gait Based Person Re-identification

Duncanson¹,

Thwaites²,

Booth³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…For example, Kale et al applied a hidden Markov model to human body contours to classify humans [5]. Derltka M and Bogdan M ensembled the kNN classifiers for human gait recognition and reached the 97.3% accuracy [42]. Sharma et al applied the artificial neural networks for gait recognition and compared the performance with BPNN (Back Propagation Neural Network), that ANN performance of the recognition method depends significantly on the quality of the extracted binary silhouettes [43].…”

Section: Related Workmentioning

confidence: 99%

CJAM: Convolutional Neural Network Joint Attention Mechanism in Gait Recognition

Jia

Zhao

et al. 2021

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

Gait recognition distinguishes one individual from others according to the natural patterns of human gaits. Gait recognition is a challenging signal processing technology for biometric identification due to the ambiguity of contours and the complex feature extraction procedure. In this work, we proposed a new model -the convolutional neural network (CNN) joint attention mechanism (CJAM) -to classify the gait sequences and conduct person identification using the CASIA-A and CASIA-B gait datasets. The CNN model has the ability to extract gait features, and the attention mechanism continuously focuses on the most discriminative area to achieve person identification. We present a comprehensive transformation from gait image preprocessing to final identification. The results from 12 experiments show that the new attention model leads to a lower error rate than others. The CJAM model improved the 3D-CNN, CNN-LSTM (long short-term memory), and the simple CNN by 8.44%, 2.94% and 1.45%, respectively.

show abstract

“…While statistical methods including t-test, ANOVA and covariance analysis were used to analyze and test the coupling or grouping across subjects based on statistic measurements for various applications of kinematic and/or physiological data (Park et al, 2017), advanced techniques of functional data analysis and machine learning demonstrated more promising results in regard to high-dimensional multi-variate gait data (Park et al, 2017). In various health related studies, supervised learning methods, such as SVM, k-nearest neighbors (KNN), linear discriminative analysis (LDA), neural network (NN), were employed for predicting or classifying in between of target and control cohorts, and usually combined with dimensionality reduction approaches, such as principle component analysis (PCA) to discover information from a high-dimensional space (Deluzio & Astephen, 2007;Coffey et al, 2011;Fukuchi et al, 2011;Eskofier et al, 2012;Andrade et al, 2013;Phinyomark et al, 2014;Janidarmian et al, 2015;Derlatka & Bogdan, 2015;Tucker et al, 2015;Watari et al, 2016;Phinyomark et al, 2016;Rida et al, 2016). In addition, some researches were focused on modeling motion dynamics of multivariate kinematic data using hidden Markov model (Mannini & Sabatini, 2012) and Bayesian network (Moon & Pavlović, 2008) for gait pattern recognition.…”

Section: Pattern Recognition and Analysis Of Human Motion Datamentioning

confidence: 99%

Self-expressive subspace clustering to recognize motion dynamics for chronic ankle instability

Qian

Yen

Folmar

et al. 2019

IISE Transactions on Healthcare Systems Engineering

View full text Add to dashboard Cite

Ankle sprains and instability are major public health concerns. Up to 70% of individuals do not fully recover from single ankle sprains and eventually develop chronic ankle instability (CAI).The diagnosis of CAI has been mainly based on self-report rather than objective biomechanical measures. The goal of this study is to quantitatively recognize the motion patterns of a multi-joint coordinate system using gait data of bilateral hip, knee, and ankle joints, and further distinguish CAI from control cohorts. We propose an analytic framework, where the concept of subspace clustering is applied to characterize the dynamic gait patterns in a lower dimensional subspace from an interdependent network of multiply joints. A support vector machine model is built to validate the learned measures compared to traditional statistical measures in a leave-one-subject-out cross validation. The experimental results showed >70% classification accuracy on average for the dataset of 47 subjects (24 with CAI and 23 controls) recruited to examine in our designed experiment. It is found that CAI can be observed from other joints (e.g., hips) significantly, which reflects the fact that there exists inter-dependency in the multi-joint coordinate system. The proposed framework presents a potential to support clinical decisions using quantitative measures during diagnosis, treatment, rehabilitation of gait abnormality caused by physical injuries (e.g., ankle sprains in this study) or even central nervous system disorders.

show abstract

Ensemble kNN classifiers for human gait recognition based on ground reaction forces

Cited by 28 publications

References 27 publications

The Most Discriminant Components of Force Platform Data for Gait Based Person Re-identification

The Most Discriminant Components of Force Platform Data for Gait Based Person Re-identification

CJAM: Convolutional Neural Network Joint Attention Mechanism in Gait Recognition

Self-expressive subspace clustering to recognize motion dynamics for chronic ankle instability

Contact Info

Product

Resources

About