Final-year B.Sc. Computer Engineering student at the University of Tabriz. My current work focuses on computer vision, learning-based 3D geometry and mesh processing, and industrial image analysis.
My broader research interests include generative modeling, multimodal and vision-language models, foundation models, and medical image analysis.
- Context-Aware Neural Subdivision — Research implementation and experimentation around learning-based mesh subdivision: a reproduced Neural Subdivision-style pipeline, mesh preprocessing, correspondence validation, training/evaluation infrastructure, reproducibility tests, and an implemented, experimentally evaluated extension. Further research is ongoing. Related stylized subdivision experiments document reproduction work.
- Froth image analysis and segmentation — U-Net, DeepLab, and SAM-family pipelines for industrial froth imagery, covering model training, evaluation, and mask analysis. Experimental work completed; a data-paper manuscript is currently in preparation.
- Froth Motion Analysis — Industrial computer vision work on froth motion and stability, with image registration, local motion estimation, and visual/CSV analysis tools.
- DebtPal — A student Django project for shared-debt tracking and settlement calculation, with tests, CI, and deployment notes.
Tools: Python, PyTorch, OpenCV, NumPy, Git, LaTeX.