People

"If you have build castles in the air, your work need not be lost; that is where they should be, now put the foundations under them." - Thoreau

Dr. Albert's professional goal in life is to leverage machine learning to automate the collection and inference of clinically useful health information to improve clinical research. His projects in wearable sensor analytics have improved the measurement of health outcomes for individuals with Parkinson's disease, stroke, and transfemoral amputations with a variety of additional populations and contexts including children with cerebral palsy as well as healthy toddler activity tracking. Current projects include video-based activity tracking and mobile robotic platforms, all in an effort to improve measures of clinical outcomes to justify therapeutic interventions.

Thasina is our lab manager coming all the way from Bangladesh in 2019. Generally, she is applying machine learning strategies to more precisely measure health outcomes due to therapeutic interventions. Her first project was creating a unified measure of health outcomes for individuals with transfemoral amputations to indicate a consistent improvement in outcomes when using a microprocessor controlled knee. She is currently working on a mobile phone application to automatically segment speakers in a conversation which was originally used to moderate group conversations, but she will use the tool to quantify speaking and social engagement for people undergoing therapy, including individuals with aphasia, to improve their ability to communicate.

Steve has developed a machine learning model for sensors placed on the arm to automatically measure spasticity (a condition in certain muscles are continuously contracted causing stiffness or tightness of the muscles and can interfere with normal movement) to enable more precision in discerning spasticity treatments.

He is currently working with the Shriners Hosptials for Children on a predictive model of surgical outcomes for children with cerebral palsy, aiding surgeons in the selection of therapeutic options to improve gait.

Sahar comes to the lab with experience in natural language processing and using machine learning to extract valuable information from diverse text corpora. Her long-term interests are in using machine learning to build models of sensory neural processing and relating that knowledge to neural models of language learning. In particular, she is studying how the visual system in developing animals uses an "Innate visual learning" strategy in which spontaneous neural activity patterns train the visual system prior to eye opening in the same way the visual system adapts to information after eye opening.

Himanshu joined UNT in Fall 2018 and completed his Master’s in Computer Science (Spring 2020). Now he has joined Computer Science PhD program in Fall 2020. He is currently working on EKG signal compression using deep neural network autoencoders. His interest includes computational neuroscience and natural language processing. He has interest in cognitive computing, natural language processing and computational robotics.

Chandrashekhar joined UNT in 2020 currently pursuing his PhD with specialization in Machine learning and Deep learning. His is currently working on scoring symptoms of Parkinson's Disease through measurements during quiet standing. He is currently involved in applying the predictive models to clinical decision support. He is interested in application of Artificial Intelligence in clinical care.

Saba joined the lab as a PhD Student in Fall 2020, and is pursuing combined Kalman filter and deep learning strategies to improve markerless tumor tracking to improve radiation oncology outcomes.

Riyad has been working as a remote student at Chittagong University of Engineering and Technology and will be formally joining the lab in Spring 2021. He has a strong interest in developing validated machine learning models, collaborating with the Biomedical AI lab and submitting a paper to ITiCSE noting high experiences and interests in proper machine learning model validation. Continuing his work in automated sign language recognition, he will continue to collaborate on projects related to gesture recognition for use in medical applications, particular for individuals unable to speak.

Chengping came to UNT in 2018 and is engaged in Masters thesis work in reinforcement learning. He is the project leader in this research effort with three Texas Academy of Math and Science students, creating and studying the behavior of a system that learns both how to play arbitrary games (tactics) and how to optimally engage opponents for maximum rewards while learning (strategy).

He received his MBA from Missouri State University, and received his BS in Information Systems from Fuzhou University (福州大学).

Theo is an Artificial Intelligence graduate student who joined the lab in 2020. He is currently working on Stock2vec Vector Embeddings using deep learning neural networks and dimensionality reduction techniques. He leads a team of two Texas Academy Math and Science students researching vector representations of stocks to predict various business outcomes. He has done some research work in recommender systems and also the application of NLP methods in intellectual properties.

Ryan is pursuing his masters in artificial intelligence with a concentration in biomedical engineering. He is interested in computational neuroscience and image processing. His current work in the lab is in applying efficient coding techniques and ICA to an Android app. The app will demonstrate the receptive field filters, similar to what is observed in our own brains, that are associated with user given or predefined natural images and sounds.

Akansha joined the lab in Spring 2020 and led the research effort creating a visual dashboard system to reduce excessive sound exposure during music instruction. Currently, she is a member of the ECG vest team using predictive modeling to identify features of cardiac arrythmias in an ECG vest created by EE faculty and students at UNT.

Namratha regularly participates in projects through the adjoining Human Intelligence and Language Technologies (HiLT) Lab. She is pursuing a computational neuroscience project demonstrating how the early visual and auditory systems can be understood through efficient coding - in essence you can "derive" a visual system from a (neurally) appropriate efficient coding of natural scenes. She has created a Jupyter notebook demonstrating this in a number of modalities (black and white, color, and "natural" audio input) and presented at SfN 2019 (the Society for Neuroscience conference), with a goal of making the notebook readily accessible to anyone, computational or not, that is interested in understanding a link between computer science and neuroscience.

Past students