cv
Overview of my academic and professional background. For more details download the PDF version.
Basics
Name | Gürkan Soykan |
Title | AI Researcher, Software Engineer |
grkansoykan@gmail.com | |
Url | https://www.linkedin.com/in/gurkan-soykan |
Summary | I am an AI scientist and researcher with a strong computer science and natural sciences foundation. Currently pursuing a PhD at Wageningen University, focusing on adaptive digital twins for precision agriculture. |
Work
-
2024.07 - Present Ph.D. Candidate
Wageningen University, Information Technology Group
Engaged in developing Adaptive Digital Twins for Precision Agriculture, integrating sensor data with AI techniques such as deep learning and reinforcement learning.
-
2020.09 - 2024.07 AI Researcher
KUIS AI Center, Koç University
During my masters: Developed various AI projects in the domain of comics, focusing on multimodal ai and for a year focused on multlingual NLP
- ComicBERT
- Spatially Augmented Comic-MTL
- Comics Text+
- Identity-Aware Comic Char. Re-Id.
- Self-Supervised Comic Face Generation
- Multi-modal Emotion Recognition
- Multilingual Instruction Tuning
-
2019.09 - 2020.05 iOS Developer & Backend Engineer
Carla Car Rental
Maintained the iOS application and led the development of rebranding releases, integrating APIs from 20+ car rental suppliers into the backend stack.
-
2018.09 - 2019.09 iOS Developer
NaylaLabs
Contributed to the development of 8 iOS applications, ranging from social networking to child development apps.
-
2018.02 - Present Software Developer
Freelance Developer
Worked on various projects including mobile applications and backend development. Notable projects include Blinq, Gökçeada tour guide, and financial chart UI for Fingaroo.
Education
-
2024.07 - Present The Netherlands
-
2023.09 - 2024.07 Istanbul, Turkey
-
2020.09 - 2023.06 Istanbul, Turkey
-
2013.09 - 2018.06 Istanbul, Turkey
-
2012.09 - 2017.06 Istanbul, Turkey
Projects
- 2021.01 - 2023.06
ComicVerse: Expanding the Frontiers of AI in Comic Books with Holistic Understanding
Master's thesis that consists of four parts to create holistic pipeline (from comic page to neural representation of the panels) for contextual comic understanding
- A Comprehensive Gold Standard and Benchmark for Comics Text Detection and Recognition
- Spatially Augmented Speech Bubble to Character Association via Comic Multi-Task Learning
- Identity-Aware Semi-Supervised Learning for Comic Character Re-Identification
- ComicBERT: A Transformer Model and Pre-training Strategy for Contextual Understanding in Comics
- 2021.01 - 2021.06
Self-Supervised Face Generation using Panel Context Information
Developed a model to generate masked faces of characters in sequential comics frames using GANs and VAEs.
- 2021.10 - 2022.01
Multimodal Emotion Recognition on Drawings
Improved VisualBERT to recognize emotions using visual and linguistic cues in comic scenes.
Skills
Programming Languages | |
Python | |
JavaScript | |
TypeScript | |
Swift | |
Objective-C | |
Kotlin | |
C++ | |
Julia |
Machine Learning Frameworks | |
PyTorch | |
TensorFlow | |
Knet |
Mobile Development | |
iOS Development | |
Flutter |
Languages
Turkish | |
Native speaker |
English | |
Fluent |
Interests
Artificial Intelligence | |
Machine Learning | |
Deep Learning | |
Natural Language Processing |
Software Engineering | |
Backend Development | |
Mobile Development |