Cyberiad Lab · Koç University

Researching machines that understand and create visual worlds

We develop models that connect perception, language, motion, and generation. Our work combines foundational machine learning with benchmarks and systems designed to reveal how visual intelligence works—and where it still fails.

01 Generative modeling Controllable image and video generation, editing, motion planning, and diffusion. Explore this area

02 Multimodal reasoning Grounding, compositional generalization, and evaluation across vision and language. Explore this area

03 Structured video Continuous, compact representations for coherent motion, reconstruction, and editing. Explore this area

04 Spherical visual intelligence Geometry-aware 360° perception, visual attention, eye tracking, and spatial audio. Explore this area

How we work

Models, evidence, and useful systems

Structure

We build representations that expose objects, motion, time, geometry, and language.

Evaluation

We design benchmarks that test specific capabilities instead of relying only on aggregate scores.

Translation

We connect foundational ideas to practical systems for media, sensing, and human-centered computing.

Join us

We welcome MSc and PhD students interested in generative AI, multimodal reasoning, computer vision, graphics, and next-generation visual systems. See our open positions and meet the research team.

Funding and support

Our research has received support from national research programs and industry partners.

Principal investigator

  • TÜBİTAK 1001 — The Support Program for Scientific and Technological Research Projects, Award #120E501, 2021–2024 · project page
  • TÜBİTAK 1003 — Primary Subjects R&D Funding Program, Award #116E685, 2017–2020 · project page
  • TÜBİTAK 3501 — Career Development Program, Award #113E497, 2014–2017 · project page

Co-investigator

  • TÜBİTAK 1007 — Public Institutions Research Funding Program, Award #114G028, 2016–2019 · project page
  • TÜBİTAK 1001, Award #113E116, and the European Union COST Programme, ICT COST IC1037 Action, 2014–2017 · project page
  • TÜBİTAK 3501 — Career Development Program, Award #112E146, 2012–2015 · project page

Industry grants and donations

  • NVIDIA Academic Grant, 2026
  • Adobe Research Gift, 2025, 2023, and 2020
  • NVIDIA Hardware Donation Grant (Tesla K40 GPU), 2016