Projects

Region-Aware Segmentation-Oriented Super-Resolution (RA-SOSR)

Investigates super-resolution-assisted segmentation methods for medical MRI and CT images. Rather than treating super-resolution as a pure image-quality step, the work studies how region-aware super-resolution, which focuses on foreground and boundary regions, relates to downstream segmentation accuracy.

M.Sc. thesis at Hacettepe University, advised by Assoc. Prof. Aydın Kaya.

• Python
• PyTorch
• Deep Learning
• Medical Image Segmentation
• Computer Vision

TARİF: Multimodal Matching and Enrichment of DIY Videos

Undergraduate research fellow on a TÜBİTAK 1001 project that matches and enriches do-it-yourself videos with a multimodal approach. Focused on recipe videos: aligning each instruction with the part of the video where it happens, using content-based text analysis, visual–textual matching and visual concept learning.

• Python
• Multimodal Learning
• Computer Vision
• NLP

Zeybek Hukuk Danışmanlık Website

A professional, fully responsive website for a legal consultancy firm, featuring a dynamic admin panel, service listings and contact forms. The backend is built with Spring Boot and PostgreSQL; the frontend uses React and Tailwind CSS.

• React
• Spring Boot
• PostgreSQL
• Tailwind CSS