A Wearable Activity Recognition Device Using Air-Pressure and IMU Sensors

Yang, Di; Huang, Jian; Tu, Xikai; Ding, Guangzheng; Shen, Tong; Xiao, Xiling

doi:10.1109/access.2018.2890004

Cited by 34 publications

(16 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…HAR aims to identify and classify the activity conducted by the agent using a model with minimum error. Conventional machine learning methods, such as support vector machine [12], random forest [13] and logistic regression [14] require hand-crafted features from the dataset. This is a challenge for sensory-based datasets because of the large size of the dataset and presence of noise in the data.…”

Section: Related Workmentioning

confidence: 99%

Deep ConvLSTM Network with Dataset Resampling for Upper Body Activity Recognition Using Minimal Number of IMU Sensors

2021

View full text Add to dashboard Cite

Human activity recognition (HAR) is the study of the identification of specific human movement and action based on images, accelerometer data and inertia measurement unit (IMU) sensors. In the sensor based HAR application, most of the researchers used many IMU sensors to get an accurate HAR classification. The use of many IMU sensors not only limits the deployment phase but also increase the difficulty and discomfort for users. As reported in the literature, the original model used 19 sensor data consisting of accelerometers and IMU sensors. The imbalanced class distribution is another challenge to the recognition of human activity in real-life. This is a real-life scenario, and the classifier may predict some of the imbalanced classes with very high accuracy. When a model is trained using an imbalanced dataset, it can degrade model’s performance. In this paper, two approaches, namely resampling and multiclass focal loss, were used to address the imbalanced dataset. The resampling method was used to reconstruct the imbalanced class distribution of the IMU sensor dataset prior to model development and learning using the cross-entropy loss function. A deep ConvLSTM network with a minimal number of IMU sensor data was used to develop the upper-body HAR model. On the other hand, the multiclass focal loss function was used in the HAR model and classified minority classes without the need to resample the imbalanced dataset. Based on the experiments results, the developed HAR model using a cross-entropy loss function and reconstructed dataset achieved a good performance of 0.91 in the model accuracy and F1-score. The HAR model with a multiclass focal loss function and imbalanced dataset has a slightly lower model accuracy and F1-score in both 1% difference from the resampling method. In conclusion, the upper body HAR model using a minimal number of IMU sensors and proper handling of imbalanced class distribution by the resampling method is useful for the assessment of home-based rehabilitation involving activities of daily living.

show abstract

Section: Related Workmentioning

confidence: 99%

Deep ConvLSTM Network with Dataset Resampling for Upper Body Activity Recognition Using Minimal Number of IMU Sensors

2021

View full text Add to dashboard Cite

show abstract

“…Moreover, the system based on air pressure has the characteristics of safety and flexibility, which is widely used in human interaction systems [ 26 , 27 ]. Yang et al [ 28 ] has proved that the HAR system’s accuracy will be improved when muscle motion data is added to motion information such as attitude angle and acceleration. In our study, we developed a compact wearable system that incorporates an inertial measurement unit (IMU) module and air pressure module.…”

Section: Introductionmentioning

confidence: 99%

Personalized Human Activity Recognition Based on Integrated Wearable Sensor and Transfer Learning

Wang

et al. 2021

Sensors

Self Cite

View full text Add to dashboard Cite

Human activity recognition (HAR) based on the wearable device has attracted more attention from researchers with sensor technology development in recent years. However, personalized HAR requires high accuracy of recognition, while maintaining the model’s generalization capability is a major challenge in this field. This paper designed a compact wireless wearable sensor node, which combines an air pressure sensor and inertial measurement unit (IMU) to provide multi-modal information for HAR model training. To solve personalized recognition of user activities, we propose a new transfer learning algorithm, which is a joint probability domain adaptive method with improved pseudo-labels (IPL-JPDA). This method adds the improved pseudo-label strategy to the JPDA algorithm to avoid cumulative errors due to inaccurate initial pseudo-labels. In order to verify our equipment and method, we use the newly designed sensor node to collect seven daily activities of 7 subjects. Nine different HAR models are trained by traditional machine learning and transfer learning methods. The experimental results show that the multi-modal data improve the accuracy of the HAR system. The IPL-JPDA algorithm proposed in this paper has the best performance among five HAR models, and the average recognition accuracy of different subjects is 93.2%.

show abstract

“…The advent of micro electro mechanical systems (MEMS) technology has enabled the development of inertial measurement units (IMU) which due to their low cost, light weight and reduced power consumption, are being widely used in a variety of different applications which require attitude (pitch and roll) and/or heading (yaw) information such as human activity and gesture recognition using smart phones and watches [1]- [3], motion stabilization and control using drones and robots [4], [5], space and marine vehicle navigation [6], 3D motion tracking systems [7] and indoor localization [8], [9].…”

Section: Introductionmentioning

confidence: 99%

“…The measurement noise covariance(M t ) is adaptively increased during dynamic conditions.The external acceleration is modeled as first order low pass filtered white noise process, The M t is a function of predicted external acceleration, and. = y a t − c a a t−1 , M t = σ3 A I 3 + acc…”

mentioning

confidence: 99%

Cascaded Kalman Filtering-Based Attitude and Gyro Bias Estimation With Efficient Compensation of External Accelerations

2020

View full text Add to dashboard Cite

We consider the problem of attitude estimation of rigid bodies in motion using low cost inertial measurement unit (IMU). An efficient scheme is proposed using two different Kalman filters by deriving their measurement models for precise attitude (pitch and roll) estimation in the presence of high and prolonged dynamic conditions and gyro bias. Both filters work in a coupled fashion where one of the filters provides accurate estimates of rigid body attitude and external acceleration using the accelerometer in conjunction with the gyroscope while the second filter is responsible to estimate the gyro bias, allowing the proposed scheme to be used in any application with minimal calibration. A new threshold based external acceleration detection module is also introduced to change the confidence level on external acceleration prediction to assist the estimation process. The proposed scheme is tested and compared with other existing estimators in the literature under different dynamical conditions and real-world experimental data sets in order to validate its effectiveness. INDEX TERMS Attitude estimation, gyro bias, external acceleration, MEMS IMU, Kalman filter.

show abstract

A Wearable Activity Recognition Device Using Air-Pressure and IMU Sensors

Cited by 34 publications

References 46 publications

Deep ConvLSTM Network with Dataset Resampling for Upper Body Activity Recognition Using Minimal Number of IMU Sensors

Deep ConvLSTM Network with Dataset Resampling for Upper Body Activity Recognition Using Minimal Number of IMU Sensors

Personalized Human Activity Recognition Based on Integrated Wearable Sensor and Transfer Learning

Cascaded Kalman Filtering-Based Attitude and Gyro Bias Estimation With Efficient Compensation of External Accelerations

Contact Info

Product

Resources

About