Enbiya Çabuk

Control and Automation Engineering student at ITU building machine learning systems at the intersection of control theory and deep learning. Designed a multi-headed CNN architecture fusing time-series and image modalities for control system identification. Built a deep learning framework from scratch with custom autograd, convolutional layers, and transformer blocks. Current work includes viewpoint-invariant object detection using LoRA-adapted vision transformers with 3D-consistent synthetic data.