Gizem Unlu
Google Scholar / GitHub / LinkedIn / Email

Hi! I am Gizem Unlu, PhD candidate at University College London (UCL) in the Computer Science Department. I am supervised by Prof. Gabriel Brostow.

During my PhD, I focused on AI, Computer Vision, and Human Computer Interaction. My research is about 3D reconstruction and modeling and generative AI, specifically converting images into 3D models using deep learning. My work is also interactive; I am passionate about building AI-powered super tools to increase human productivity and creativity.

Publications

3D Modeling of Multi-Building Scenes from Prototyping to Vertex-Efficient Representations

Gizem Esra Unlu, Mohamed Sayed, Yulia Gryaditskaya, and Gabriel Brostow

Under Review 2026

Preprint Paper

GroundUp: Rapid Sketch-Based 3D City Modeling

Gizem Esra Unlu, Mohamed Sayed, Yulia Gryaditskaya, and Gabriel Brostow

European Conference on Computer Vision (ECCV) 2024

Project Page, Paper(Arxiv), Video,

Interactive Sketching of Mannequin Poses

Gizem Esra Unlu, Mohamed Sayed, Gabriel Brostow

International Conference on 3D Vision (3DV) 2022

Project Page, Paper, Video, Bibtex

Person Inpainting with Generative Adversarial Networks

Gizem Esra Unlu

European Conference on Computer Vision Workshop (ECCVW) 2018

Paper,
Education

PhD Computer Vision (University College London)

Supervised by Prof. Gabriel Brostow and Dr. Iasonas Kokkinos.

MSc Computer Engineering (Bogazici University)

Thesis: "Image Deblurring from Sign Language Videos" Supervised by Prof. Lale Akarun.

Bachelor of Computer Engineering (Istanbul Technical University) (Double Major)

Thesis: 3D Facial Reconstruction from RGB Images. Supervised by Prof. Hazim Ekenel.

Bachelor of Mathematics Engineering (Istanbul Technical University) (Double Major)

Thesis: Filtering Techniques in Speckle Noise Reduction. Supervised by Dr. Burcu Tunga.

Work Experience

2025
Grenoble, FR

Autodesk Research, Research Scientist Internship

HCI & Visualization Team
Built a multimodal LLM-based UI understanding/automation agent.
Enabling real-time UI interpretation and task guidance.

2024
Grenoble, FR

Naver Labs Europe, Research Scientist Internship

3D Human Understanding Team
Built an interactive multi-person 3D pose annotation tool.

Awards

October 2023 -

June 2024

PhD Enrichment Scheme Placement award

The Alan Turing Institute

September 2018

ECCV 2018 Looking at People Satellite Workshop Challenge

Image Completion Competition 1st place

Publication: Person Inpainting with Generative Adversarial Networks

Teaching

University College London

Machine Vision TA - 2020, 2021, 2022, 2023

Bogazici University

Computer Vision TA - 2019

Skills

Languages

Python, C++, MATLAB, Java/Javascript

Deep Learning Libraries

Pytorch(primary), Tensorflow

Useful Tools

Blender, Adobe Illustrator, Adobe After Effects