M. Arda Aydın

M. Arda Aydın
arda.aydin@bilkent.edu.tr

I am a master’s student at Bilkent University, working with Prof. Tolga Çukur in the Imaging and Computational Neuroscience (ICON) Lab.

My research focuses on medical image analysis. I study how vision-language and foundation models can be efficiently adapted to medical imaging tasks, with a particular interest in segmentation and understanding how these models make their predictions.

Previously, I worked with Prof. Gözde Ünal at the ITU Vision Lab on training-free open-vocabulary semantic segmentation.

You can find more about my work on my Google Scholar page and CV.

Email

News

2026

Our preprint ACE-LoRA is available on arXiv.

2025

Presented ITACLIP at CVPR 2025 Workshops in Nashville.

Selected Publications

ACE-LoRA method overview
arXiv

ACE-LoRA: Graph-Attentive Context Enhancement for Parameter-Efficient Adaptation of Medical Vision-Language Models

M. Arda Aydın, Melih B. Yılmaz, Aykut Koç, and Tolga Çukur

arXiv preprint · 2026

Parameter-efficient adaptation of medical vision-language models to specialized domains using graph-attentive context enhancement.

ITACLIP method overview
CVPRW

ITACLIP: Boosting Training-Free Semantic Segmentation with Image, Text, and Architectural Enhancements

M. Arda Aydın, Efe M. Çırpar, Elvin Abdinli, Gözde Ünal, and Yusuf H. Şahin

IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops · 2025

Turning image-level vision-language representations into pixel-level predictions through image, text, and architectural enhancements, without additional training.

Education

2024–present

Bilkent University

M.S. in Electrical and Electronics Engineering · Ankara, Türkiye

2019–2024

Istanbul Technical University

B.S. in Electronics and Communication Engineering · Istanbul, Türkiye