Serkan Sulun
Doctoral Researcher at Faculdade de Engenharia da Universidade do Porto
Porto, Porto, Portugal
Summary
👤
Senior🎓
Top SchoolSerkan Sulun is a deep learning researcher and doctoral candidate with nine years of experience specializing in multimodal and generative AI for multimedia, currently based at INESC TEC in Porto. His PhD work produced state-of-the-art multimodal video analysis models for emotion and cinematic genre prediction and novel conditional transformers that generate symbolic music from video, outperforming prior approaches. He works across video, image, audio, MIDI and text using architectures from CNNs and transformers to diffusion models and VLMs, and routinely ships reproducible code and preprints on arXiv and GitHub. A practical engineer as well as an academic, he leverages PyTorch/TensorFlow, SLURM and AWS for large-scale experiments and has applied research to industrial safety, audio enhancement generalization, and future frame prediction. Beyond research, he designs and teaches graduate-level programming courses, aiming to make advanced models accessible to non-researchers through easy-to-run open-source releases.
9 years of coding experience
1 year of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Faculdade de Engenharia da Universidade do PortoBachelor of Science - BS, Electrical and Electronics Engineering, 3.86/4, Bachelor of Science - BS, Electrical and Electronics Engineering, 3.86/4 at Sabanci UniversityMaster of Science - MS, Electrical and Electronics Engineering, 3.81/4, Master of Science - MS, Electrical and Electronics Engineering, 3.81/4 at Koç University
English, Portuguese, Turkish
Github Skills (85)
music-generation10
midi9
nlp9
python9
mean-square-error9
pytorch9
emotion9
machine-learning9
loss9
ssim8
transformer8
audio8
weight8
deep-learning8
gpu8
Programming languages (6)
Github contributions (5)
serkansulun/midi-emotion
Apr 2022 - Dec 2022
Generates multi-instrument symbolic music (MIDI), based on user-provided emotions from valence-arousal plane.
Contributions:12 commits, 14 pushes, 5 branches in 7 months
deep-learningemotionmidimusic-generationsymbolic-music
serkansulun/serkansulun.github.io
Mar 2020 - Mar 2025
Contributions:1 PR, 119 pushes, 2 branches in 5 years 1 month
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.