Deep Transfer Learning Based Human Activity Recognition By Transforming IMU Data To Image Domain Using Novel Activity Image Creation Method

Human Activity Recognition is the most popular research area in the pervasive computing field in recent years. Sensor data plays a vital role in identifying several human actions. Convolutional Neural Networks (CNNs) have now become the most recent technique in the computer vision phenomenon, but still it is premature to use CNN for sensor data, particularly in ubiquitous and wearable computing. In this paper, we have proposed the idea of transforming the raw accelerometer and gyroscope sensor data to the visual domain by using our novel activity image creation method (NAICM). Pre-trained CNN (AlexNet) has been used on the converted image domain information. The proposed method is evaluated on several online available human activity recognition dataset. The results show that the proposed novel activity image creation method (NAICM) has successfully created the activity images with a classification accuracy of 98.36% using pre trained CNN.

Detecting Coronary Artery Disease Using Rest Seismocardiography and Gyrocardiography

In this study, we present a non-invasive solution to identify patients with coronary artery disease (CAD) defined as ⩾50% stenosis in at least one coronary artery. The solution is based on the analysis of linear acceleration (seismocardiogram, SCG) and angular velocity (gyrocardiogram, GCG) of the heart recorded in the x, y, and z directional axes from an accelerometer/gyroscope sensor mounted on the sternum. The database was collected from 310 individuals through a multicenter study. The time-frequency features extracted from each SCG and GCG data channel were fed to a one-dimensional Convolutional Neural Network (1D CNN) to train six separate classifiers. The results from different classifiers were later fused to estimate the CAD risk for each participant. The predicted CAD risk was validated against related results from angiography. The SCG z and SCG y classifiers showed better performance relative to the other models (p < 0.05) with the area under the curve (AUC) of 91%. The sensitivity range for CAD detection was 92–94% for the SCG models and 73–87% for the GCG models. Based on our findings, the SCG models achieved better performance in predicting the CAD risk compared to the GCG models; the model based on the combination of all SCG and GCG classifiers did not achieve higher performance relative to the other models. Moreover, these findings showed that the performance of the proposed 3-axial SCG/GCG solution based on recordings obtained during rest was comparable, or better than stress ECG. These data may indicate that 3-axial SCG/GCG could be used as a portable at-home CAD screening tool.

A Revolutionary Gaming Style in Motion

From the timeline of the year, 2012 MONECT has been aiming towards the conceptuality for developing the formulation of making a virtual remote controller for a wide range of variety within the context considering various types of devices and peripherals consisting within the prospective realm of virtual controlling. Moving forward, where recently in the timeline for the year of 2017, the including of the functionality of that very same aspect with numerous advancements which was termed and computed as a remote desktop session with gaming control for a wide variety of games which includes games like Racing, Frames Per Seconds (FPS), Role-Playing Game (RPG) along with many more where each type of gaming aspect was equipped with its own perspective type of setup and a familiar type layout for the users who were considered for having different types of controllers for each specific gaming style and associated gameplay render. The project prospect evolved further within the year timeline of 2019–2021 which introduced and revolved around the rapidly deployable features and functionality with integrated advancements in terms of computing and gaming as a whole. Based on that deployment project outcome and developmental scope of the research, the application utilized the full use of the provided onboard sensors to give the user the ultimate experience while performing gameplay (for example, like the Accelerometer sensor, G-Sensor, Gyroscope sensor, Camera sensor etc. with many more). Each of the sensors controlled a different particular aspect of control. For instance, Frames Per Second (FPS) mode triggered and enabled the Gyroscope sensor which would allow the user to aim at their perspective targets for a solid headshot kill. On the other hand, the Race mode used the G-Sensor to enable steering mode of movement in the form of any vehicle. Besides that, the virtual remote sessions brought about the privilege and also gave each user a simultaneous interaction among devices and peripherals with real-time remote access at any given moment in time of usage.

The application of smartphone sensors to promote cognitive abilities easier and more effective physics learning

The use of sensor on smartphones as physics learning media encourages teachers to reconstruct teaching methods. This paper presents the effect on students’ cognitive abilities using an accelerometer and a gyroscope sensor simultaneously in learning circular motion, as well as students’ responses and the effectiveness of sensor media on smartphones used. A pre-experimental research design was used in this study which involved 12 students of XII MIPA at SMA Negeri 1 Pagaden Subang Jawa Barat. They learned circular motion guided by the smartphone sensors-based worksheets and were tested using an essay test for cognitive abilities measurements. Meanwhile, students’ responses and the effectiveness were obtained using a Likert scale questionnaire. The improvement of students’ cognitive abilities was significantly higher than the pretest which was obtained from the N-gain with a final value of 0.41. Worksheets using smartphone sensor media were more effective than conventional learning. In addition, students showed a positive response in which 93.75% of students were interested, 85,42% were motivated to learn the circular motion, and 91.14% of students became easier in understanding physics concepts.

Deep Transfer Learning Based Human Activity Recognition By Transforming IMU Data To Image Domain Using Novel Activity Image Creation Method

Human Activity Recognition is the most popular research area in the pervasive computing field in recent years. Sensor data plays a vital role in identifying several human actions. Convolutional Neural Networks (CNNs) have now become the most recent technique in the computer vision phenomenon, but still it is premature to use CNN for sensor data, particularly in ubiquitous and wearable computing. In this paper, we have proposed the idea of transforming the raw accelerometer and gyroscope sensor data to the visual domain by using our novel activity image creation method (NAICM). Pre-trained CNN (AlexNet) has been used on the converted image domain information. The proposed method is evaluated on several online available human activity recognition dataset. The results show that the proposed novel activity image creation method (NAICM) has successfully created the activity images with a classification accuracy of 98.36% using pre trained CNN.

WR-Hand

This paper presents WR-Hand, a wearable-based system tracking 3D hand pose of 14 hand skeleton points over time using Electromyography (EMG) and gyroscope sensor data from commercial armband. This system provides a significant leap in wearable sensing and enables new application potentials in medical care, human-computer interaction, etc. A challenge is the armband EMG sensors inevitably collect mixed EMG signals from multiple forearm muscles because of the fixed sensor positions on the device, while prior bio-medical models for hand pose tracking are built on isolated EMG signal inputs from isolated forearm spots for different muscles. In this paper, we leverage the recent success of neural networks to enhance the existing bio-medical model using the armband's EMG data and visualize our design to understand why our solution is effective. Moreover, we propose solutions to place the constructed hand pose reliably in a global coordinate system, and address two practical issues by providing a general plug-and-play version for new users without training and compensating for the position difference in how users wear their armbands. We implement a prototype using different commercial armbands, which is lightweight to execute on user's phone in real-time. Extensive evaluation shows the efficacy of the WR-Hand design.

Roll Angle Estimation Algorithm of Geomagnetic/Gyro Combination Based on an Interacting Multiple-Model Kalman Filter

The knowledge of the geomagnetic and gyro information that can be used for projectile roll angle is decisive to apply trajectory correction and control law. In order to improve the measurement accuracy of projectile roll angle, an interacting multiple-model Kalman filter (IMMKF) algorithm using gyro angular rate information to geomagnetic sensor information is proposed. Firstly, the data acquisition module of the geomagnetic sensor and the gyroscope sensor is designed, and the test data of the sensors are obtained through the semiphysical experiments. Furthermore, according to the measurement accuracy of each sensor, the algorithm performs the IMMKF process on the geomagnetic/gyro information to get the roll angle. It can be proven by experiments and calculation results that the error of the roll angle obtained after processing by the IMMKF algorithm is close to 2°, which is better than the 5° calculated by adopting the Kalman filter directly with geomagnetic information.

OPTIMALISASI PENERAPAN TEKNOLOGI VIRTUAL REALITY TOUR KAMPUS UNIVERSITAS MAJALENGKA MENGGUNAKAN METODE MULTIMEDIA DEVELOPMENT LIFE CYCLE

Campus introduction media in general still use billboards and brochures or advertisements, therefore the use of Virtual Reality technology is very good to use. Optimizing the application of this technology is made to build an interactive application that can visualize the facilities and infrastructure at University Majalengka in the form of 360 images and videos to make it easier for users to digitally view the facilities and infrastructure of the University Majalengka Campus. The method used in this research is MDLC (Multimedia Development Life Cycle). This method can optimize the manufacturing process and application of technology used in a smartphone media that is integrated with other media or 3D media. MDLC rule pattern itself by prioritizing interactive rests on the user. This application was built using the C # programming language with Unity 3D software. The final result of this research is the creation of a virtual tour application of the University Majalengka Campus which can be used by users to view interactive digital facilities and infrastructure on campus with a 360 degree angle using a smartphone that has a Gyroscope sensor which can be integrated with the paired virtual reality technology. into virtual reality glasses.

Validity and reliability of a novel monitoring sensor for the quantification of the hitting load in tennis

Wearable sensor systems are a emerging tools for the evaluation of the sport’s activity and can be used to quantify the external workload of the athlete. The main goal of this paper was to evaluate the validity and reliability of the “Armbeep inertial measurement unit” (IMU) sensor both in a closed tennis exercise and in open matchplay. Twentyfour junior tennis players performed a baseline drill and played matches, during which they wore a combined accelerometer and gyroscope sensor. Video footage was concomitantly recorded using a digital video camera. The agreement between the measurements was assessed with the intraclass correlation coefficient (ICC) and the standard error of measurement (SEM). A simple linear regression was used to predict the number of shots registered from the video and from the Armbeep IMU sensor’s data. The number of total forehand and backhand shots during the drill repetitions showed an excellent test and re-test reproducibility (ICC≥0.90). There was a significant relationship between the Armbeep IMU sensor’s number of contacts and the total number of shots (R2 = 0.938) which indicated the excellent reliability of the tested Armbeep IMU sensor for those parameters. Considering the accuracy of the total tennis shots and the small magnitude of error for wrist speed and acceleration, the Armbeep IMU sensor appears to be an appropriate on-court tool that can be used to monitor the hitting load during tennis practice and matches.

Design of Tripod Auto Balance With Gyroscope Sensor

Dalam dunia fotografi tentunya tidak asing dari peralatan pendukung seperti tripod yang merupakan alat untuk membantu agar badan kamera bisa berdiri dengan tegak dan tegar. Dalam menggunakan tripod kamera tentunya harus mengatur kaki-kaki tripod secara manual agar tripod menjadi seimbang dan sejajar. Oleh karena itu penulis bertujuan membuat inovasi terhadap tripod yang secara otomatis menyesuaikan kaki-kakinya agar tripod dapat berdiri secara seimbang dengan memanfaatkan sensor gyroscope. Penelitian ini menggunakan metode pengembangan atau Research and Development (R&D). Adapun sumber data yang diperoleh adalah data-data primer dan sekunder dari proses realisasi. Data tersebut dakan dianalisa dengan metode interaktif. Selanjutnya perancangan sistem yang terdiri dari sensor MPU 6050 sebagai sensor pembaca kemiringan tripod. Untuk penggerak kaki tripod digunakanlah rangkaian motor stepper yang arah putaran dan kecepatannya diatur oleh driver A4988. Sebagai pengendali mikrokontroler, dipilih Arduino nano yang memakai pemrograman Bahasa C dengan menggunakan metode cut on-off sebagai penerapan dalam mencari keseimbangan tripod. Berdasarkan analisa di atas menghasilkan kesimpulan bahwa tripod dapat seimbang secara otomatis dengan memanfaatkan sensor gyroscope dan dalam berbagai permukaan, tripod mampu bertahan seimbang dengan kemiringan sudut maksimal 10°.