Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430462
L. G. Pagani, L. Guerinoni, L. Falorni, P. Fedeli, G. Langfelder
The work presents a consistent follow-up on a study to identify the critical impact energy of suspended polysilicon masses onto stoppers [1], to prevent creation of debris and particles which affect the lifetime stability of MEMS inertial sensors. The proposed test device is capable of detecting debris generation by monitoring the leakage current between buried electrodes underneath the mass-stopper contact region. Results indicate that leakage begins to appear at 15 nJ impact energy, and manifests on 100% of tested structures at energies of 25 nJ. Optical analyses after device uncapping confirm the prediction, validating the test method.
{"title":"Finding the critical impact energy for micro debris generation in MEMS inertial sensors","authors":"L. G. Pagani, L. Guerinoni, L. Falorni, P. Fedeli, G. Langfelder","doi":"10.1109/INERTIAL51137.2021.9430462","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430462","url":null,"abstract":"The work presents a consistent follow-up on a study to identify the critical impact energy of suspended polysilicon masses onto stoppers [1], to prevent creation of debris and particles which affect the lifetime stability of MEMS inertial sensors. The proposed test device is capable of detecting debris generation by monitoring the leakage current between buried electrodes underneath the mass-stopper contact region. Results indicate that leakage begins to appear at 15 nJ impact energy, and manifests on 100% of tested structures at energies of 25 nJ. Optical analyses after device uncapping confirm the prediction, validating the test method.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123395869","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430484
Ujjval N. Patel, Imraan A. Faruque
The growing availability of low-cost commercial inertial measurement units (IMUs) raises questions about how to best improve sensor estimates when using multiple IMUs. This paper reports on the performance of two approaches applied to GPS-denied onboard attitude estimation. The approaches are a virtual IMU approach fusing sensor measurements and a Federated Filter fusing state estimates from several Extended Kalman Filters (EKFs) each using one IMU and magnetometer. We compare their performance as quantified by root mean square (RMS) using parallel implementations of estimators in a Raspberry-Pi-based autopilot during prescribed motions in a motion capture volume. The results suggest that a Multi-IMU GPS-denied approach can deliver comparable performance to the single-IMU GPS aided approach and provide a testbed for multi-IMU performance quantification.11Portions of this work received support from NASA University Leadership Initiative grant 80NSSC20M0162.
{"title":"Sensor Fusion To Improve State Estimate Accuracy Using Multiple Inertial Measurement Units","authors":"Ujjval N. Patel, Imraan A. Faruque","doi":"10.1109/INERTIAL51137.2021.9430484","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430484","url":null,"abstract":"The growing availability of low-cost commercial inertial measurement units (IMUs) raises questions about how to best improve sensor estimates when using multiple IMUs. This paper reports on the performance of two approaches applied to GPS-denied onboard attitude estimation. The approaches are a virtual IMU approach fusing sensor measurements and a Federated Filter fusing state estimates from several Extended Kalman Filters (EKFs) each using one IMU and magnetometer. We compare their performance as quantified by root mean square (RMS) using parallel implementations of estimators in a Raspberry-Pi-based autopilot during prescribed motions in a motion capture volume. The results suggest that a Multi-IMU GPS-denied approach can deliver comparable performance to the single-IMU GPS aided approach and provide a testbed for multi-IMU performance quantification.11Portions of this work received support from NASA University Leadership Initiative grant 80NSSC20M0162.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"310 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123091270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430489
O. Halevy, Stella Lulinsky, S. Krylov
We report on a novel architecture of a resonant accelerometer incorporating four proof masses and a compliant parallel motion linkage as a force amplifier. A simple, Manhattan geometry, manufacturable, device is distinguished by low parasitic compliance and purely axial, lacking any bending, loading of the sensing beams. Silicon on insulator (SOI) devices were operated in open loop and in a non-differential mode, the acceleration was emulated by an electrostatic force. Consistently with the model prediction a sensitivity of ≈ 2.3 Hz/V2, which is equivalent to ≈ 417 Hz/g, was experimentally demonstrated in a Si device with ≈ 500 µm × 480 µm × 25 µm masses and 250 µm long and ≈ 1.5 µm wide resonant sensing beams.
{"title":"Resonant Accelerometer with Compliant Parallel Motion Linkage Force Amplification Mechanism","authors":"O. Halevy, Stella Lulinsky, S. Krylov","doi":"10.1109/INERTIAL51137.2021.9430489","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430489","url":null,"abstract":"We report on a novel architecture of a resonant accelerometer incorporating four proof masses and a compliant parallel motion linkage as a force amplifier. A simple, Manhattan geometry, manufacturable, device is distinguished by low parasitic compliance and purely axial, lacking any bending, loading of the sensing beams. Silicon on insulator (SOI) devices were operated in open loop and in a non-differential mode, the acceleration was emulated by an electrostatic force. Consistently with the model prediction a sensitivity of ≈ 2.3 Hz/V2, which is equivalent to ≈ 417 Hz/g, was experimentally demonstrated in a Si device with ≈ 500 µm × 480 µm × 25 µm masses and 250 µm long and ≈ 1.5 µm wide resonant sensing beams.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130719541","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430491
T. Tsukamoto, Shuji Tanaka
In this paper, a theoretical aspect of frequency and damping mismatch compensation in a MEMS rate integrating gyroscope (RIG) controlled by CW and CCW rotational modes is reported. The mismatches of a resonator could be compensated by the amplitudes and phases of the driving signals, even if the principle axes of stiffness and damping are not aligned to the X-Y coordinate. The proposed theoretical formula well consistent with the previously reported experimental results, as well as the numerical simulation.
{"title":"Theoretical Consideration of Mismatch Compensation for MEMS Resonator Having Unaligned Principle Axes","authors":"T. Tsukamoto, Shuji Tanaka","doi":"10.1109/INERTIAL51137.2021.9430491","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430491","url":null,"abstract":"In this paper, a theoretical aspect of frequency and damping mismatch compensation in a MEMS rate integrating gyroscope (RIG) controlled by CW and CCW rotational modes is reported. The mismatches of a resonator could be compensated by the amplitudes and phases of the driving signals, even if the principle axes of stiffness and damping are not aligned to the X-Y coordinate. The proposed theoretical formula well consistent with the previously reported experimental results, as well as the numerical simulation.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124310324","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430478
J. Marconi, Giacomo Bonaccorsi, D. Giannini, L. Falorni, F. Braghin
This paper describes how mechanical nonlinearities can be exploited to obtain a frequency-matched MEMS gyroscope. Exploiting the hardening behavior of the oscillator, we show how it is possible to match drive and sense frequency by changing the drive displacement amplitude. This way, both the resonance amplitudes of the drive and sense axes are exploited, boosting the sensitivity of the device. Moreover, the near-flat drive frequency response increases both the robustness and bandwidth. A prototype of a yaw gyroscope was also manufactured to test the feasibility of the proposed approach.
{"title":"Exploiting Nonlinearities for Frequency-Matched MEMS Gyroscopes Tuning","authors":"J. Marconi, Giacomo Bonaccorsi, D. Giannini, L. Falorni, F. Braghin","doi":"10.1109/INERTIAL51137.2021.9430478","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430478","url":null,"abstract":"This paper describes how mechanical nonlinearities can be exploited to obtain a frequency-matched MEMS gyroscope. Exploiting the hardening behavior of the oscillator, we show how it is possible to match drive and sense frequency by changing the drive displacement amplitude. This way, both the resonance amplitudes of the drive and sense axes are exploited, boosting the sensitivity of the device. Moreover, the near-flat drive frequency response increases both the robustness and bandwidth. A prototype of a yaw gyroscope was also manufactured to test the feasibility of the proposed approach.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116896742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430483
I. Prikhodko, J. Geen, Carey Merritt, Sam Zhang
The paper describes a small size (2.25 mm2active sensor area) yaw rate sensor component for ADIS1654x series Inertial Measurement Units (IMUs) with notable performance parameters such as in-run bias instability of 0.55 °/hr stable over 3 months on Allan Deviation, g-sensitivity of 0.8 (°/hr)/g, bias repeatability of 50 °/hr in 10 years, and no observable startup drift from a cold-start. The current generation IMU offers improvement of bias improvement by 4x, g-sensitivity by 30x, and repeatability by 10x over previous generation ADI's IMU.
{"title":"Vibration Immune, Long-Term Stable and Low Noise Synchronized Mass MEMS Gyroscope","authors":"I. Prikhodko, J. Geen, Carey Merritt, Sam Zhang","doi":"10.1109/INERTIAL51137.2021.9430483","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430483","url":null,"abstract":"The paper describes a small size (2.25 mm2active sensor area) yaw rate sensor component for ADIS1654x series Inertial Measurement Units (IMUs) with notable performance parameters such as in-run bias instability of 0.55 °/hr stable over 3 months on Allan Deviation, g-sensitivity of 0.8 (°/hr)/g, bias repeatability of 50 °/hr in 10 years, and no observable startup drift from a cold-start. The current generation IMU offers improvement of bias improvement by 4x, g-sensitivity by 30x, and repeatability by 10x over previous generation ADI's IMU.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"437 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122478163","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430482
A. R. Anwary, Damla Arifoglu, Michael Jones, M. Vassallo, H. Bouchachia
Gait assessment relies on clinical tools based on observation by trained staffs who give a subjective opinion. Objective gait analysis via motion capture systems (e.g. Qualisys) have limited availability as they are laboratory based and require complex equipment. A low-cost user-friendly Inertial Measurement Units (IMUs) embedded insole and an Android App based personalized gait analysis system is developed for uses in home or clinics. Accelerometer and gyroscope synchronous data are collected from both right and left legs for 10 young and 10 older adults a period of 100 consecutive days. We propose an automatic gait features extraction method, real-time visualization and age-groups classification. Accuracy of stride detection method is 100% for young. Accuracy for older adults is 91% for right and 88% for left leg. Convolutional neural networks (CNNs) are used to extract features from gait data and are combined with long short-term memory (LSTM) to exploit the time information between features. This is evaluated empirically using traditional classification and deep learning techniques (CNN+LSTM RNN) regardless of feature engineering. Accuracy to classify young and older adults with CNN-LSTM, NB, SVM and J48 is 100%. Our insole-based gait analysis automatically interprets the gait features and users can monitor their gait at home using our simple visualization tool that allows widespread home-based diagnosis and management of gait abnormalities and rehabilitation.
{"title":"Insole-based Real-time Gait Analysis: Feature Extraction and Classification","authors":"A. R. Anwary, Damla Arifoglu, Michael Jones, M. Vassallo, H. Bouchachia","doi":"10.1109/INERTIAL51137.2021.9430482","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430482","url":null,"abstract":"Gait assessment relies on clinical tools based on observation by trained staffs who give a subjective opinion. Objective gait analysis via motion capture systems (e.g. Qualisys) have limited availability as they are laboratory based and require complex equipment. A low-cost user-friendly Inertial Measurement Units (IMUs) embedded insole and an Android App based personalized gait analysis system is developed for uses in home or clinics. Accelerometer and gyroscope synchronous data are collected from both right and left legs for 10 young and 10 older adults a period of 100 consecutive days. We propose an automatic gait features extraction method, real-time visualization and age-groups classification. Accuracy of stride detection method is 100% for young. Accuracy for older adults is 91% for right and 88% for left leg. Convolutional neural networks (CNNs) are used to extract features from gait data and are combined with long short-term memory (LSTM) to exploit the time information between features. This is evaluated empirically using traditional classification and deep learning techniques (CNN+LSTM RNN) regardless of feature engineering. Accuracy to classify young and older adults with CNN-LSTM, NB, SVM and J48 is 100%. Our insole-based gait analysis automatically interprets the gait features and users can monitor their gait at home using our simple visualization tool that allows widespread home-based diagnosis and management of gait abnormalities and rehabilitation.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122574686","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430479
D. Vatanparvar, A. Shkel
In Coriolis Vibratory Rate Integrating Gyroscopes (CVRIG), accuracy of angle measurement is known to be coupled to symmetry of the mechanical structure. This paper provides a study on the effect of asymmetries in control electronics on operation and accuracy of direct angle measurements. We demonstrated that gain mismatch in detection electronics affects the estimation of the pendulum variables in the CVRIG mathematical model. An error in the pendulum variables was shown to adversely affect the estimated orientation of the orbital trajectory and the closed-loop control. In the case of gain mismatch in actuation electronics, the control forces were observed to interfere with free precession of the oscillation pattern causing additional errors in the angle measurement. We proposed a method to distinguish the angle errors due to mechanical asymmetries from the angle errors caused by imperfections in control electronics. Using the method, we identified gain mismatches in the control electronics and subsequently used the identified parameters for calibration of a micro-fabricated gyroscope. By applying the method of calibration to a Dual Foucault Pendulum (DFP) gyroscope, we were able to reduce the angle bias error by 10-times and reached a 0.06 degree of precession accuracy at the input angular rate of 500 dps, without any compensation for mechanical asymmetries.
{"title":"Identification of Gain Mismatches in Control Electronics of Rate Integrating CVGs","authors":"D. Vatanparvar, A. Shkel","doi":"10.1109/INERTIAL51137.2021.9430479","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430479","url":null,"abstract":"In Coriolis Vibratory Rate Integrating Gyroscopes (CVRIG), accuracy of angle measurement is known to be coupled to symmetry of the mechanical structure. This paper provides a study on the effect of asymmetries in control electronics on operation and accuracy of direct angle measurements. We demonstrated that gain mismatch in detection electronics affects the estimation of the pendulum variables in the CVRIG mathematical model. An error in the pendulum variables was shown to adversely affect the estimated orientation of the orbital trajectory and the closed-loop control. In the case of gain mismatch in actuation electronics, the control forces were observed to interfere with free precession of the oscillation pattern causing additional errors in the angle measurement. We proposed a method to distinguish the angle errors due to mechanical asymmetries from the angle errors caused by imperfections in control electronics. Using the method, we identified gain mismatches in the control electronics and subsequently used the identified parameters for calibration of a micro-fabricated gyroscope. By applying the method of calibration to a Dual Foucault Pendulum (DFP) gyroscope, we were able to reduce the angle bias error by 10-times and reached a 0.06 degree of precession accuracy at the input angular rate of 500 dps, without any compensation for mechanical asymmetries.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115748007","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430471
Danmeng Wang, A. Shkel
This paper reports a study on contribution of electromechanical amplitude modulation (EAM) on performance of MEMS Coriolis vibratory gyroscopes (CVGs). We theoretically predicted and experimentally demonstrated the impact of EAM on both quality factor and output noise in MEMS CVGs operating in the open-loop rate mode. We demonstrated the effect on a dynamically amplified dual-mass gyroscope (DAG) improving the gyroscope performance from 0.04 deg/rt-hr in Angular Random Walk (ARW) and 0.52 deg/hr in bias instability to 0.0065 deg/rt-hr in ARW and 0.08 deg/hr in bias instability by changed the EAM setting from 3.5 to 1.2 V in amplitude and from 75 to 225 kHz in frequency. The optimized EAM parameters were derived by the noise prediction model presented in the paper.
{"title":"Effect of EAM on Quality Factor and Noise in MEMS Vibratory Gyroscopes","authors":"Danmeng Wang, A. Shkel","doi":"10.1109/INERTIAL51137.2021.9430471","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430471","url":null,"abstract":"This paper reports a study on contribution of electromechanical amplitude modulation (EAM) on performance of MEMS Coriolis vibratory gyroscopes (CVGs). We theoretically predicted and experimentally demonstrated the impact of EAM on both quality factor and output noise in MEMS CVGs operating in the open-loop rate mode. We demonstrated the effect on a dynamically amplified dual-mass gyroscope (DAG) improving the gyroscope performance from 0.04 deg/rt-hr in Angular Random Walk (ARW) and 0.52 deg/hr in bias instability to 0.0065 deg/rt-hr in ARW and 0.08 deg/hr in bias instability by changed the EAM setting from 3.5 to 1.2 V in amplitude and from 75 to 225 kHz in frequency. The optimized EAM parameters were derived by the noise prediction model presented in the paper.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"32 5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133963969","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-03-22DOI: 10.1109/INERTIAL51137.2021.9430460
Sonia Perez-Gamboa, Qingquan Sun, Yan Zhang
Non-intrusive sensor based human activity recognition (HAR) is utilized in a spectrum of applications including fitness tracking devices, gaming, health care monitoring, and smartphone applications. In this paper, we design a multi-layer hybrid architecture with Convolutional Neural Networks (CNN) and Long Short Term Memory (LSTM). Based on the exploration of a variety of multi-layer combinations, we present a lightweight, hybrid, and multi-layer model which can improve the recognition performance by integrating local features and scale-invariant with dependencies of activities. The experimental results demonstrate the efficacy of the proposed model which can achieve a 94.7% activity recognition rate on a benchmark dataset. This model outperforms traditional machine learning and other deep learning methods. Additionally, our implementation achieves a balance between accuracy and efficiency.
{"title":"Improved Sensor Based Human Activity Recognition via Hybrid Convolutional and Recurrent Neural Networks","authors":"Sonia Perez-Gamboa, Qingquan Sun, Yan Zhang","doi":"10.1109/INERTIAL51137.2021.9430460","DOIUrl":"https://doi.org/10.1109/INERTIAL51137.2021.9430460","url":null,"abstract":"Non-intrusive sensor based human activity recognition (HAR) is utilized in a spectrum of applications including fitness tracking devices, gaming, health care monitoring, and smartphone applications. In this paper, we design a multi-layer hybrid architecture with Convolutional Neural Networks (CNN) and Long Short Term Memory (LSTM). Based on the exploration of a variety of multi-layer combinations, we present a lightweight, hybrid, and multi-layer model which can improve the recognition performance by integrating local features and scale-invariant with dependencies of activities. The experimental results demonstrate the efficacy of the proposed model which can achieve a 94.7% activity recognition rate on a benchmark dataset. This model outperforms traditional machine learning and other deep learning methods. Additionally, our implementation achieves a balance between accuracy and efficiency.","PeriodicalId":424028,"journal":{"name":"2021 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134467231","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：