Humera Sabir

Graduate Research Assistant, UTA — Fulbright PhD Scholar

Research focus: Graph Representation Learning · Heterogeneous Knowledge Graphs · Scalable GNNs

Advisor: Prof. Ashraf Aboulnaga
Fulbright PhD Scholar Gold Medalist (BS-CS) 5+ years Freelancing Featured in USEFP Newsletter
Email: hxs7303 [at] mavs.uta.edu  |  sabirhumera95 [at] gmail.com
Download CV Email Humera LinkedIn GitHub

About

I am a Graduate Research Assistant in the Large-Scale Intelligent Data Systems LIDS Laboratory at the University of Texas at Arlington and a Fulbright PhD Scholar. My work centers on graph representation learning and building scalable GNNs for large, heterogeneous knowledge graphs. Previously, I served as a Lecturer at FAST (NUCES), Islamabad.

Experience

Graduate Research Assistant

University of Texas at Arlington — LIDS Lab
Aug 2023 – Present
  • Research on scalable, relation-aware graph learning for link prediction and node classification.
  • Mini-batch training on heterogeneous KGs; efficient sampling/partitioning for large graphs.

Lecturer

FAST (NUCES), Islamabad — Faculty profile
Jan 2019 – Aug 2023
  • Taught Parallel & Distributed Computing and Fundamentals of Data Science; supervised OOP, Web, ML/AI labs.

Freelance Software/ML Engineer

Remote
2018 – Present (5+ years)
  • Delivered full-stack and ML projects across web, CV, and data pipelines; client-facing delivery and maintenance.

Software Developer

CASE, Islamabad
Sep 2017 – Jan 2019
  • Cross-platform web/mobile apps including patient record systems; Angular, Express, MongoDB, AWS.

Education

PhD — Computer Science

University of Texas at Arlington
2023 – Present

MS — Data Science

FAST-NUCES, Islamabad
2019 – 2021 · CGPA 3.70/4.00

BS — Computer Science

CASE, Islamabad
2013 – 2017 · CGPA 3.81/4.00 (Gold Medalist)

Skills

Core
Python PyTorch Graph Representation Learning PyTorch Geometric Heterogeneous KGs Scalable GNNs
Data & Dev
NumPy/Pandas scikit-learn OpenCV Angular · Node MongoDB/SQL

Selected Projects

Physiological Parameters via PPG (MS Thesis)

Non-invasive estimation of hemoglobin from fingertip video using Python, OpenCV, and PyTorch.

Ridesharing Path Update

Inserting new nodes into existing worker routes; path optimization under constraints.

DBLP Graph Analytics

Co-authorship graphs and search UI from DBLP XML with Python + SQLite + Angular.

Adaptive E-commerce Recommender

Session-aware recommendations with CF and real-time feedback on a price-comparison site.

Publications

  1. Fingertip Video Dataset for Non-Invasive Diagnosis of Anemia Using ResNet-18 Classifier . H. Sabir et al., IEEE Access, vol. 12, pp. 68880–68892, 2024.
    DOI: 10.1109/ACCESS.2024.3398353
Keywords:
PPG signals Non-invasive Hemoglobin estimation Computer vision Deep learning ResNet-18 Fingertip video dataset Image processing
Show citation
H. Sabir et al., “Fingertip Video Dataset for Non-Invasive Diagnosis of Anemia Using ResNet-18 Classifier,” IEEE Access, vol. 12, pp. 68880–68892, 2024. doi: 10.1109/ACCESS.2024.3398353

Contact

For collaborations or questions, reach me at hxs7303@mavs.uta.edu or sabirhumera95@gmail.com.

I’m open to research discussions and mentoring.