CV
Research Statement
My work lies at the intersection of computer vision and natural language processing, with primary focus on multimodal learning and vision-language models. I develop diverse, semantically aligned representations for cross-modal retrieval across image, text, video, and audio using multi-embedding strategies to capture non-literal and abstract relationships. I work on structured knowledge and grounded reasoning for interpretable vision-language models, supported by new benchmarks and evaluation pipelines. This research has led to publications at top-tier venues such as ACL 2025.
Education
- Virginia Tech, Ph.D. Computer Science (Jan 2023 – Jan 2028, expected)
Advisor: Dr. Chris Thomas | GPA: 4.00/4.00 - Jordan University of Science and Technology, M.S. Data Science (Feb 2020 – Jun 2022)
Advisor: Prof. Rehab Duwairi | GPA: 4.26/4.30 - Jordan University of Science and Technology, B.S. Computer Science (Feb 2016 – Jan 2020)
Advisor: Dr. Malak Abdullah | GPA: 4.04/4.20 (Ranked 1st of 62)
Publications
See the publications page for the complete list and links.
Research Experience
- Research Assistant, Virginia Tech (Mar 2023 – Present)
- Led diversity-aware image-text retrieval using maximal matching objectives and multi-embedding strategies, improving performance by up to 7.1% and resulting in an ACL 2025 publication.
- Built a diversification pipeline that generates facet-specific captions with vision-language models, expanding beyond COCO and Flickr30k coverage.
- Leading generative room-geometry work that conditions on room-impulse responses (RIRs) to reconstruct layouts from line-segment representations.
- Collaborated with Columbia University on ENTER and JourneyBench, contributing modeling, annotation, and evaluation for NeurIPS 2024 publications.
- Supported hospital deployment of a real-time, multi-view surgical instrument classifier covering 95 instrument types with high accuracy and SUS 81.7.
- Research Assistant, Jordan University of Science and Technology (Feb 2020 – Nov 2022)
- Created So2al-wa-Gwab, a 10K-question Arabic QA dataset with scripted scraping, annotation workflow design, and three-stage quality control.
- Benchmarked span-extraction models across seven Arabic QA datasets using EM and F1, demonstrating transformer gains over baselines.
- Led an Arabic humor detection project with end-to-end NLP pipelines and ranked 7/45 (top ~15%) in shared-task competition results.
- Conducted a systematic literature review on Arabic QA extraction to highlight translation dependencies and inform dataset design.
- Research Assistant (Intern), Jordan University of Science and Technology (Jun 2019 – Aug 2019)
- Built and debugged deep learning pipelines with TensorFlow and spaCy for propaganda and emotion detection, ranking 7/26 (top 27%) in the NLP4IF 2019 SLC track.
Academic Contributions
- Graduate Teaching Assistant: CS4804 Intro to AI (Spring 2025); CS5914 AI Tools in Software Development (Fall 2023, Spring 2024, Fall 2024); CS1064 Intro to Programming in Python (Summer I 2023, Summer I 2024); CS5824/ECE5424 Advanced Machine Learning (Spring 2023).
- Reviewer: WACV 2024; ACL Rolling Review (Feb 2025, May 2025).
Honors & Awards
- Sanghani Center Travel Fund – $750, Virginia Tech (2025)
- CS Department Travel Fund – $750, Virginia Tech (2025)
- GPSS Travel Fund Grant – $500, Virginia Tech Graduate School (2025)
- Outstanding Graduate Student Award – full scholarship for M.S., JUST (2020-2022)
- Outstanding Undergraduate Student Award – ranked #1 of 62, multiple Honors List mentions (2020)
Conferences & Presentations
- ACL 2025 – 63rd Annual Meeting of the Association for Computational Linguistics, Vienna, Austria (Jul 2025)
Poster: “Maximal Matching Matters: Preventing Representation Collapse for Robust Cross-Modal Retrieval.”
Technical Skills
- Programming: Python, C++, Java, Bash, LaTeX, CSS, SQL
- ML Frameworks: PyTorch, TensorFlow, Transformers, scikit-learn, Keras
- Data Handling: NumPy, Pandas, spaCy, NLTK, seaborn, matplotlib
- Tools: Conda, Git, Weights & Biases, Docker
- Languages: English, Arabic