Matrix-free Second-order Optimization of Gaussian Splats with Residual Sampling
International Conference on 3D Vision, 2026
Doctoral Researcher · Open to collaborations
Hi, I'mHamza Pehlivan
My research focuses on efficient 3D/4D representations for real-world mapping, with current work on fast optimizers for Gaussian Splatting training.
I am a Ph.D. student in the Visual Computing and Artificial Intelligence department at the Max Planck Institute for Informatics, advised by Prof. Dr. Christian Theobalt. Before joining MPI, I completed my M.Sc. at Bilkent University with Asst. Prof. Aysegül Dündar in Bilkent Generative DLR Lab, working on real image editing and reconstruction with StyleGAN.
Publications
Warping the Residuals for Image Editing with StyleGAN
International Journal of Computer Vision, 2024
StyleRes: Transforming the Residuals for Real Image Editing with StyleGAN
IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023
Diverse Inpainting and Editing with GAN Inversion
International Conference on Computer Vision, 2023
Benchmarking the Robustness of Instance Segmentation Models
IEEE Transactions on Neural Networks and Learning Systems, 2022
Teaching
- Apr 2024 – Jul 2024
Seminar Supervisor · Saarbrücken, Germany
Saarland University
- Supervised the “Volume Rendering” topic in the seminar Classical Concepts of Computer Vision and Computer Graphics in the Neural Age.
- Sep 2021 – Jun 2023
Teaching Assistant · Ankara, Turkey
Bilkent University
- Computer Networks (3 semesters): designed projects on TCP/UDP protocols.
- Introduction to Python (1 semester): led weekly labs and supported students in programming work.
- Mentored undergraduate research students at Bilkent Generative DLR Lab with Docker and GPU-server tutorials.
Experience
- Sep 2021 – Jun 2023
Research Assistant · Ankara, Turkey
Bilkent Generative Deep Learning Research Lab
- Researched high-fidelity image reconstruction, image editing, and inpainting using StyleGAN; advised by Asst. Prof. Aysegül Dündar.
- Contributed to facial attribute editing methods compared against StyleGAN-based baselines.
- Analyzed robustness of instance segmentation models under lighting, weather, and noise perturbations.
- Jun 2021 – Jul 2021
Summer Intern · Ankara, Turkey
Jotform
- Built a website-based recommendation system: crawled and clustered websites into topics with Python and scikit-learn.
- Jun 2020 – Jul 2020
Summer Intern · Ankara, Turkey
A2 Technology
- Worked on an under-vehicle inspection system combining classical vision and deep learning.
- Implemented image comparison with SIFT descriptors using OpenCV and C++.
Education
- Nov 2023 – present
Ph.D. in Computer Science · Saarbrücken, Germany
Saarland University & MPI for Informatics
- Doctoral researcher in the Visual Computing and Artificial Intelligence department.
- Sep 2021 – Jun 2023
M.Sc. in Computer Science · Ankara, Turkey
Bilkent University
- Full departmental scholarship for academic achievement during undergraduate studies.
- CGPA: 3.96 / 4.00.
- Sep 2016 – Jun 2021
B.Sc. in Computer Science · Ankara, Turkey
Bilkent University
- Comprehensive scholarship awarded for high ranking (top ~0.07%) in the University Placement Exam.
- CGPA: 3.52 / 4.00.
Skills
Programming
Frameworks & Tools
Languages
Contact
Open to research discussions, collaborations, and paper questions.
Email is the best way to reach me. I read every message and reply when I can.