At a glance
The Intelligent Multimodal Computing and Sensing (IMICS) team is part of the Computer science department – Texas State University.
The research goals of the IMICS team are the development of theoretical and technological solutions in the area of Machine Learning, Data Mining and Computer Vision focus in applications of Smart Health and Well-being, Assisted Living and Bioinformatics.
The research goals of the IMICS team are the development of theoretical and technological solutions in the area of Machine Learning, Data Mining and Computer Vision focus in applications of Smart Health and Well-being, Assisted Living and Bioinformatics.
Major Projects
Biosignal Analysis
Bio-signal analysis project was part of larger project of developing VR environment for veteran who are suffering from PTST to help returning to their normal lives more gracefully.
My role was to create a method to collect bio-signal data while participants are experiencing stressors (crowded environments in VR, questions which might cause stress in interview). I successfully collected bio-signals which included ECG, GSR, Pulse oximeter, Thermal Imaging, Respiratory rate from over 40 participants.
After collecting and cleaning the data I also tried to apply different machine learning approaches to the data including traditional machine learning techniques with feature extraction and deep learning using TensorFlow.
My role was to create a method to collect bio-signal data while participants are experiencing stressors (crowded environments in VR, questions which might cause stress in interview). I successfully collected bio-signals which included ECG, GSR, Pulse oximeter, Thermal Imaging, Respiratory rate from over 40 participants.
After collecting and cleaning the data I also tried to apply different machine learning approaches to the data including traditional machine learning techniques with feature extraction and deep learning using TensorFlow.
Kinect data collection utility
Kinect data recorder was initially made to do motion analysis of Kinect streams. I build the application for Kinect V2 and can and it can records Color, RGB & Skeleton data at about 28fps (depending on hardware configuration).
Later the project was extended to do motion analysis using dynamic time warping algorithm as well as for movement classification.
Later the project was extended to do motion analysis using dynamic time warping algorithm as well as for movement classification.
