Deep neural network approach for estimating the three-dimensional human center of mass using joint angles

Chebel, Elie; Tunç, Burcu

doi:10.1016/j.jbiomech.2021.110648

Cited by 7 publications

(15 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The predicted foot pressure is further combined with image-based foot localization to calculate BoS and CoP. These three stability components (CoM, CoP, and BoS) are combined to calculate CoMNet predicts image-based 3D CoM from 3D pose with a mean Euclidean error of 17.56 mm, outperforming the state of the art method using body worn inertial sensors [30], and predicting an error nearly as low as the expected error in ground truth motion capture calculations [45] while using only image-based data. Additionally, the work originally published in [10] reporting CoP and BoS results for one take per subject sub-sampling is validated here for all valid dataset performances (Figures 5 and 6), confirming the sub-sampling in [10] was a representative cross section.…”

Section: Discussionmentioning

confidence: 99%

“…Additionally, all image-based configurations produce similar and consistent results. Using only image-based pose input, CoMNet establishes a state-of-the-art better than the mean Euclidean error of 18.1 mm achieved by Chebel et al [30], which requires subject measurements and inertial sensors.…”

Section: A Com Predictionmentioning

confidence: 92%

“…Chebel et al [30] present a state of the art neural network for 3D CoM estimation of stationary subjects using two subject height measurements and 11 inertial sensors measuring joint angles as input. They report 3D RMSE errors in a subject-centric format that generates a mean Euclidean error of 18.1mm when using a full body model tested on new subjects.…”

Section: A Stabilitymentioning

confidence: 99%

See 2 more Smart Citations

Image-based Stability Quantification

Scott¹,

Challis²,

Collins³

et al. 2022

Preprint

View full text Add to dashboard Cite

Quantitative evaluation of human stability using foot pressure/force measurement hardware and motion capture (mocap) technology is expensive, time consuming, and restricted to the laboratory (lab-based). We propose a novel imagebased method to estimate three key components for stability computation: Center of Mass (CoM), Base of Support (BoS), and Center of Pressure (CoP). Furthermore, we quantitatively validate our image-based methods for computing two classic stability measures against the ones generated directly from labbased sensory output (ground truth) using a publicly available multi-modality (mocap, foot pressure, 2-view videos), tensubject human motion dataset. Using leave-one-subject-out cross validation, our experimental results show: 1) our CoM estimation method (CoMNet) consistently outperforms state-of-theart inertial sensor-based CoM estimation techniques; 2) our image-based method combined with insole foot-pressure alone produces consistent and statistically significant correlation with ground truth stability measures (CoMtoCoP R=0.79 P<0.001, CoMtoBoS R=0.75 P<0.001); 3) our fully image-based stability metric estimation produces consistent, positive, and statistically significant correlation on the two stability metrics (CoMtoCoP R=0.31 P<0.001, CoMtoBoS R=0.22 P<0.001). Our study provides promising quantitative evidence for stability computations and monitoring in natural environments.Index Terms-image-based, stability, base of support, center of mass, center of pressure, deep learning.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: A Com Predictionmentioning

confidence: 92%

Section: A Stabilitymentioning

confidence: 99%

See 1 more Smart Citation

Image-based Stability Quantification

Scott¹,

Challis²,

Collins³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Recent techniques have been introduced to enhance the estimation process, most notably those which relied on machine learning 22 – 26 . In our previous work 27 , we introduced a method to map the joint angles to the CoM position using a deep neural network (DNN). In that study, the CoM used in the training dataset was collected from ‘Fit subjects’ using the segmental analysis method, which can be less accurate in cases where the subjects have different body densities as in the case of obesity.…”

Section: Introductionmentioning

confidence: 99%

Evaluation of center of mass estimation for obese using statically equivalent serial chain

Chebel

Tunç

2022

Sci Rep

Self Cite

View full text Add to dashboard Cite

The complex structure of the human body makes its center of mass (CoM) estimation very challenging. The typically used estimation methods usually suffer from large estimation errors when applied to bodies with structural differences. Thus, a reliable estimation method is of utmost importance. In this paper, we present a detailed evaluation of a subject-specific CoM estimation technique named Statically Equivalent Serial Chain (SESC) by investigating its estimation ability over two different groups of subjects (Fit and Obese) in comparison to the segmental analysis method. For this study, we used an IMU-based motion capture system and a force platform to record the joint angles and corresponding center of pressure (CoP) values of twenty-five participants while performing a series of static postures. The root-mean-square errors (RMSE) of SESC’s estimation for both groups showed close and lower mean values, whereas the segmental analysis method showed significantly larger RMSE values in comparison to SESC (p < 0.05). In addition, we used the Bland–Altman analysis to evaluate the agreement between the two techniques and the ground truth CoP, which showed the accuracy, precision, and reliability of SESC over both groups. In contrast, the segmental analysis method did not present neither accurate nor precise estimations, as our analysis revealed considerable fixed and proportional biases.

show abstract

“…temperature, density) can influence SBPADs' results. MoCap technologies (Asadi & Arjmand, 2020;Chebel & Tunc, 2021) are also able to be used for posture evaluations, providing digitalization of the subjects' motion. Regarding posture strain and muscular fatigue evaluation, the most used method is EMG (Mudiyanselage, Nguyen, Rajabi & Akhavian, 2021), a technique based on the measurement of skin's electrical potential through the use of electrodes.…”

Section: Introductionmentioning

confidence: 99%

Endüstri̇ 4.0 Dönüşümünde İnsan Faktörleri̇ Mühendi̇sli̇ği̇: Iot Temelli̇ Teknoloji̇ler Analizi̇

Kaya

2022

Endüstri Mühendisliği

View full text Add to dashboard Cite

Using advanced technologies and devices in human factors engineering (HFE) processes is becoming a rising trend in international arena, regarding Industry 4.0 philosophy and transformation consummation. Transition to this new technology from traditional HFE applications offers many advantages but also refers to the analysis of a very complex set of numerous emerging criteria conflicting in varying directions and dimensions. This study focuses on that enigma and investigates the problem space to facilitate Ergonomics 4.0 transformation process with the employment of fuzzy sets theory, Delphi method and Best-Worst Method (BWM). New technologies and devices introduced within Industry 4.0 era for instrument based ergonomic assessment, occupational health and safety applications, and, physical environment monitoring were addressed as another contribution of this study to Ergonomics 4.0 aspect. An evaluation framework apropos of related challenging decision structures was proposed in the wake of in-depth literature analysis, where, the validated criteria set was clarified with fuzzy Delphi Method. The elucidated criteria list was than observed with BWM to propose a transition period charter. Main and sub-criteria of the problem were scrutinized according to decision hierarchy; local and global importance levels of criteria, and, outcomes regarding different parties of the decision making process were interpreted comparatively in details, and suggestions has been made in the light of multi-dimensional benchmarking debates.

show abstract

Deep neural network approach for estimating the three-dimensional human center of mass using joint angles

Cited by 7 publications

References 22 publications

Image-based Stability Quantification

Image-based Stability Quantification

Evaluation of center of mass estimation for obese using statically equivalent serial chain

Endüstri̇ 4.0 Dönüşümünde İnsan Faktörleri̇ Mühendi̇sli̇ği̇: Iot Temelli̇ Teknoloji̇ler Analizi̇

Contact Info

Product

Resources

About