College of Computer Science and Mathematics

Discussion of a PhD thesis in the College of Computer Science and Mathematics – Department of Computer Science entitled:
(An Android-based Smartphone Model for Human Activity Recognition Using Deep Learning Approach)
A doctoral thesis was discussed at the College of Computer Science and Mathematics at the University of Mosul on Wednesday, 22-6-2022, tagged:
(An Android-based Smartphone Model for Human Activity Recognition Using Deep Learning Approach)
For the student, Adiba Khaled Abbou, and under the supervision of Prof. Dr. Laheeb Mohamed Ibrahim
The thesis submitted by the student dealt with a smartphone model to identify human activity using smart technologies, focusing on detecting the individual movement of a person using sensor data.The study dealt with the use of deep learning models to successfully accomplish human activity recognition tasks that included (Convolution Neural Network (CNN), Principal Component Analysis (PCA), and Long-Short-Term-Memory (LSTM). ).
Activities are identified individually (CNN, LSTM) or in a hybrid form ((CNN-LSTM), (PCA-CNN), (PCA-LSTM), (PCA-CNN-LSTM)). Additionally, the Time-LeNet(T-leNet) model was developed to create a new, more efficient Modified-Time-LeNet (M-T-leNet) model. To ensure that the models perform adequately in the data processing, they were tested in a public database (UCI HAR) in the Offline System and the models that performed best in the Offline System were applied in the private smartphone application Our Online System, which represents the models (“CNN-LSTM with an accuracy of 96.709% and M-T-leNet with an accuracy of 90.058%”).The study aims to develop a real-time activity recognition system for Android phones. Using three types of sensors to track the activity of the elderly: accelerometers, linear accelerometers, and gyroscope sensors. To track one of the six activities (“lying, sitting, standing, walking, going upstairs, and walking downstairs”) data is collected from sensors and sent to a smartphone; The smartphone app reads the sensor values ​​and uses them as input for the proposed model.
The discussion committee consisted of:The discussion committee consisted of:
Prof. Duha Bashir Abdullah / University of Mosul / College of Computer Science and Mathematics / President
Prof. Dr. Fawzia Mahmoud Ramo / University of Mosul / College of Computer Science and Mathematics / member
Prof. Naglaa Akram Younes / University of Mosul / College of Computer Science and Mathematics / Member
Prof. Dr. Ahmed Talib Abdul-Amir/ Middle Technical University/ Administrative Technical College/ Member
Prof. Dr. Adel Sabri Issa/ Zakho University/ College of Science/ Member
Prof. Dr. Laheeb Muhammad Ibrahim/Mosul University/College of Computer Science and Mathematics/member and supervisor