Thasina Tabashum, Ph.D. Candidate
email: thasinatabashum@my.unt.edu Lab: F206, UNT Discovery Park, 3940 North Elm Street, Denton, TX 76207-7102. Google Scholar | Linkedin | GithubEducation:
Ph.D. Candidate, Computer Science, and Engineering, University of North Texas, Fall’19 - Present, 4.00/4.00. Supervisor: Dr. Mark Albert
M.Sc, Artificial Intelligence - machine learning, University of North Texas, Fall’19 - May'23, 4.00/4.00.
B.Sc, Computer Science, and Engineering, American International University-Bangladesh (2018), 3.90/4.00.
NEWS! Received TAPIA scholarship, and attending TAPIA 2023 !!
Current Research Focus:
Graduate Research Assistant, BiomedAi Lab, University of North Texas, [Fall’22 - Present]
Developing medical outcome prediction model utilizing rehabilitation clinical text. Using BERT and T5 clinical models, and exploring in-context learning methods to predict WeeFIM scores of cerebral palsy children.
Collaboration: Shriners Hospitals for Children https://www.shrinerschildrens.org/en,
Research Experiences:
Graduate Research Assistant, BiomedAi Lab, University of North Texas, [Fall’19 - Fall'21]
Research Projects:
Quantify gait data: Developing medical outcome measures using PCA/autoencoders. To quantify gait measurement for individuals with movement disorders, we propose data-driven autoencoder approaches to create summary metrics. These metrics help clinicians evaluate patients and can be used to predict outcomes of surgery, therapy, and rehabilitation.
2. Developing a gait summary metric using an autoencoder
Collaboration: Shriners Hospitals for Children https://www.shrinerschildrens.org/en, Shirley Ryan AbilityLab https://www.sralab.org/
Speaker segmentation for speech pathology applications: To evaluate speech impairment and social engagement at home, we are developing a machine learning pipeline integrated with mobile applications.
Paper: Conversation Moderator: A Mobile App for Tracking Individual Speaking in Group Conversations. | Project Link
Presentations: 1. Mobile Diarization Dashboard Application and Remote Vocalization Sensor Prototype for Evaluating Communication Rehabilitation Effectiveness.
Collaboration: Deep Sensor Information Extraction Lab
Automating pupil tracking: Developed a pupil size extraction system to assess Parkinson's Disease that will assist clinicians in developing biomarkers in a cost-effective way. The initial prototype was built using the Kalman Filter.
Paper: Detection of Parkinson's Disease Through Automated Pupil Tracking of the Post-illumination Pupillary Response | Project Link
Collaboration: Loyola University Chicago Stritch School of Medicine, Deep Sensor Information Extraction Lab
Research Student, Natural Language Processing and Machine Learning Lab, Daffodil International University, [March’18 - Dec’18]
Research Focus: Sentiment Extraction | Paper
Industry Experiences:
Deep learning Intern, Quantiphi, Boston, MA [May'22 - Aug '22]
Applying attention-based models to quantify impaired gaits using ground force reaction sensors.
Junior Software Engineer, Dynamic Solution Innovators Limited, [Jan’19 - Jul’19]
Responsibilities: Full Stack Web Developer. Technologies: Java, Hibernate, Vuejs, Jest.
Teaching Experiences:
Graduate Teaching Assistant- University of North Texas, Denton, Texas, [Aug'22 - May'23]
Courses: Bioinformatics, Advanced Bioinformatics, Introduction to Algorithms.
Graduate Teaching Assistant- University of North Texas, Denton, Texas, [Jan'22 - May'22]
Courses: Methods in Empirical Analysis.
Publications:
Tabashum T, Xiao T, Jayaraman C, Mummidisetty CK, Lipschutz R, Mathur G, Jayaraman A, Albert MV. Autoencoder Composite Scoring to Evaluate Prosthetic Performance in Individuals with Lower Limb Amputation. Smart Health,n. Bioengineering MDPI 2022.
Tabashum T, Zaffer A, Yousefzai R, Colletta K, Jost MB, Park Y, Chawla J, Gaynes B, Albert MV, Xiao T. Detection of Parkinson’s Disease through Automated Pupil Tracking of the Post-illumination Pupillary Response. -Frontier in Medicine, 2021.
S Zelman, M Dow, T Tabashum, T Xiao, MV Albert. Accelerometer-Based Automated Counting of Ten Exercises without Exercise-Specific Training or Tuning. - Journal of Healthcare Engineering, 2020.
Xiao, T., Tabashum, T., Jebamalaidass, R., Du, A., Leal, M., Oliviera, E., Metwally, B., Albert, M. (2020). Conversation Moderator: A mobile app for tracking individual speaking in group conversations. 4th Workshop on Semantic Multimedia Computing (SMC ‘20), part of the 14th IEEE International Conference on Semantic Computing (ICSC 2020), San Diego, Feb 3-5, 2020.
T Tabashum, AM Shaykat, S Abujar, M Mohibullah, S Chanda. “Sentiment Extraction From Text Using Emotion Tagged Corpus”, 2019 10th International Conference on Computing, Communication and Networking Technologies(ICCNT)
Presentations:
Tabashum T, Wang SJ, Karen Kruger K, KrzakJ, Graf A, Albert MV. “Autoencoder-derived Single Summary Metric to Assess Gait Quality”. American Congress of Rehabilitation Medicine (ACRM) conference 2021.
Tabashum T, Xiao T, Gaynes B, Chawla J, Colleatta D, Albert MV. "Automated Pupil Tracking For Parkinson's Disease Biomarker Detection By Integrating Kalman Filters In A Robust User Interface." Biomedical Engineering Society conference (BMES 2020) Oct 14-17, 2020.
Xiao T, Tabashum T, Olness G, Mahbub I, Berman D, Tasneem NT, Albert MV. “Mobile Diarization Dashboard Application and Remote Vocalization Sensor Prototype for Evaluating Communication Rehabilitation Effectiveness.” American Congress of Rehabilitation Medicine (ACRM) conference 2020.
Zelman, S., Dow, M., Tabashum, T., Albert, M., Xiao, T. (2020). Automatic counting methods applied to unspecified repetitive physical activities. ACM Richard Tapia Conference 2020.
Albert, M., Mummidisetty, C., Tabashum, T., Jayaraman, A. (2019). PCA Composite Evaluation of Outcomes for Microprocessor Knee versus Mechanical Knee in Individuals with Dysvascular Transfemoral Amputations. Biomedical Engineering Society conference (BMES 2019).
Awards:
Tapia Scholarship. | 2023, and 2020.
Computer Research Association Widening Participation (CRP-WP) GRAD COHORT for women Attendee (Scholarship recipient)| 2023.
Technical Talk, “What it's like for the Humans behind Advances in Artificial Intelligence”, Digital Diva, CSTA Dallas Fort Worth, 2021.
Grace Hopper Celebration Attendee (Building, Recruiting, and Inclusion for Diversity scholarship)| 2021, and 2020.
2nd Runner Up, National Women Hackathon, Bangladesh, 2017
2nd Runner Up, AIUB Girls Programming Contest, 2017.
Nominated for Women Leadership Summit, Le Meridien by IPDC, 2017.
YouthQuake Finalists- Youth Campaign For Earthquake Preparedness, United Nations Development Programme, 2017.
IEEE Asia-Pacific (R10) SYWL Congress – Bangalore, India. Presented IEEE American International University-Bangladesh Student Branch and received several awards for the brunch and Bangladesh section, 2016.
The Duke of Edinburgh Award, UK, Bronze, 2012.
Relevant Certifications and Courses:
Course: Machine Learning For Precision Medicine | Project: Integrating multi-omics and prior biological data to predict Colorectal cancer cells using Single Cell Trio-seq data | Collaboration: https://biocomp.engineering.unt.edu/home
Neuromatch Academy Deep Learning | https://academy.neuromatch.io
Spring ’21: Software Development for AI.
Coursera, 2020: Deep Learning, Machine Learning with TensorFlow on Google Cloud Platform.
Fall ’20: Natural Language Processing, AI for wearables.
Spring ’20: Big Data and Data Science, Deep Learning.
Fall ’19: Information Retrieval, Machine Learning.
Undergrad Project Thesis: Geo-Social Sentiment Analysis Based on Social Media Data.