Contact
- Porto, Portugal
- serkan.sulun@inesctec.pt
- github.com/serkansulun
Education
Doctorate
Electrical and Computer Engineering
University of Porto
2019
- Present
Masters
Electrical and Electronics Engineering
Koc University
2016
- 2018
GPA: 3.81
Bachelors
Electronics Engineering
Sabanci University
2009
- 2014
GPA: 3.86
Skills
Deep learning
Pytorch
Tensorflow
Generative Networks
Adversarial Learning
Recurrent Neural Networks
Convolutional Neural Networks
Transformers
Programming
Python
MATLAB
C++
Julia
Signal processing
MIDI processing
Audio processing
Image processing
Video processing
Video compression
Computer vision
Music production
Cubase
Protools
Audacity
Music21
MIDI
Miscellaneous
Remote development
Machine learning
Git
LaTeX
Ubuntu
Windows
Languages
-
English | Fluent
-
Turkish | Fluent
-
Portuguese | Intermediate
-
German | Beginner
Serkan Sulun
Deep Learning Researcher
Summary
I use deep neural networks on music and video.
Experience
Researcher
INESC TEC, Portugal
|
Mar 2019
- Present
Enhancement of complex music signals using deep neural networks
Video-based symbolic music generation
Research assistant
Koc University, Turkey
|
Sep 2016
- Nov 2018
(2 years 2 months)
Video compression and frame prediction using deep neural networks, including generative adversarial networks
Teaching assistant
Koc University, Turkey
|
Sep 2016
- Nov 2018
(2 years 2 months)
Leading laboratory sessions and occasionally lectures on programming with MATLAB
Sound engineer
Telfs Youth Center Studio, Austria
|
Oct 2014
- Aug 2015
(11 months)
Recording and mixing music performed by local youngsters
Intern
Champalimaud Neuroscience Program, Portugal
|
Jul 2013
- Oct 2013
(3 months)
Processing neural signals and videos obtained by rat experiments, designing the controller board. Article acknowledgement: ncbi.nlm.nih.gov/pmc/articles/PMC4721960
Peer instructor
Sabanci University, Turkey
|
Feb 2010
- Jun 2014
(4 years 3 months)
Lecturing in small study groups and occasionally in amphitheaters on freshman chemistry and biology
Publications
Symbolic music generation conditioned on continuous-valued emotions (2022)
IEEE Access
DOI:10.1109/ACCESS.2022.3169744
|
arXiv:2203.16165
On Filter Generalization for Music Bandwidth Extension Using Deep Neural Networks (2021)
IEEE Journal of Selected Topics in Signal Processing
DOI:10.1109/JSTSP.2020.3037485
|
arXiv:2011.07274
Can Learned Frame-Prediction Compete with Block-Motion Compensation for Video Coding? (2021)
Signal, Image and Video Processing
DOI:10.1007/s11760-020-01751-y
|
arXiv:2007.08922
Deep Learned Frame Prediction for Video Compression (2018)
Koc University, Masters Thesis
Thesis:531218
|
arXiv:1811.10946
Awards
Doctoral Fellowship
| Foundation for Science and Technology, Portugal
Doctoral Fellowship
| la Caixa Foundation
Graduate Fellowship (Masters)
| Koc University
Ranked 9 out of 204,392 in Turkish Academic Personnel and Graduate Education Exam
Excellence Merit Scholarship (Bachelors)
| Education Foundation of Turkey
Excellence Merit Scholarship (Bachelors)
| Sabanci University
Employment Scholarship
| Sabanci University
Bachelors Scholarship
| Prime Ministry of Turkey
Ranked 297 out of 1,349,782 in Turkish University Entrance Exam
Erasmus+ Exchange Grant
| European Commission
Erasmus+ Internship Grant
| European Commission
Interviews
Reviewer
International Society for Music Information Retrieval Conference
Transactions of the International Society for Music Information Retrieval
IEEE Access
Teaching experience
Introduction to deep learning (invited talk)
| Polytechnic Institute of Porto, Portugal
Deep learning for music generation (invited talk)
| University of Porto, Portugal
Deep learning for music information retrieval (invited talk)
| University of Porto, Portugal
Introduction to MATLAB
| Koc University, Turkey
Freshman mathematics, physics, chemistry and biology (peer teacher)
| Sabanci University, Turkey
Supervision
Pedro Miguel Pereira de Oliveira
Emotion classification of MIDI lyrics using transformers (Masters thesis, 2022)