Muhammed Pektaş

Muhammed Pektas

I'm a PhD student at University of Bakircay Computer Engineering department under the supervision of Dr. Murat Uçar. In addition, I'm working as a Senior Machine Learning Engineer at HubX now. We are building and growing highly-scalable mobile apps that reached over 100 million users across 160 countries here. I am focusing on customer specific content creation and related computer vision tasks such as image animation, restoration, enhancing, inpainting, text-to-image and image-to-image generation etc. I do research and development about both the generative and descriptive modeling tasks that are related to computer vision and machine learning to improve the quality and efficiency of the app. Also, I prepare the ml models for production and serving them. In addition, I have a Computer Engineering BSc. and MSc. degree. My BSc. degree from Konya Technical University advised by Prof. Erkan Ülker and MSc. degree from University of Ege advised by Prof. Aybars Uğur.

Nowadays, my main research areas are Computer Vision, Generative Modelling, and Efficient Deep Learning.

Github Linkedin Google Scholar Twitter

Publications


Performance Analysis of Efficient Deep Learning Models for Multi-Label Classification of Fundus Image
EPerformance Analysis of Efficient Deep Learning Models for Multi-Label Classification of Fundus Imag
Muhammed Pektas
AITA Github

Efficient Hair Style Transfer with Generative Adversarial Networks

Efficient Hair Style Transfer with Generative Adversarial Networks
Muhammed Pektas, Baris Gecer, Aybars Ugur
Arxiv Github

Realistic Hair Synthesis with Generative Adversarial Networks

Realistic hair synthesis with generative adversarial networks
Muhammed Pektas, Aybars Ugur - Master Thesis
Arxiv Yöktez

Open Source Projects


Identity Similarity (idsim)

This repository can help researchers who want to use face recognition in their research. BFAS - Brute Force Architecture Search
Muhammed Pektas
Github

Pysimpler

This package simplifies the fundamental software engineering practices such as bottleneck analysis, cache cleaning etc.
Muhammed Pektas
Github

BFAS - Brute Force Architecture Search

BFAS - Brute Force Architecture Search
Muhammed Pektas
Github

Highlights and News

  • March 2023 - I shared IdentitySimilarity (idsim) pypi package in Github .
  • December 2023 - I gave a speech titled "Diffusion Models : The new favorite method in Generative AI" at Google Developer Group Devfest Izmir 2023.
  • October 2023 - "Performance Analysis of Efficient Deep Learning Models for Multi-Label Classification of Fundus Image" is published at AITA.
  • August 2023 - I shared pysimpler pypi package in Github.
  • January 2023 - I started to work as Senior Machine Learning Engineer at HubX.
  • October 2022 - I shared BFAS pypi package in Github.
  • September 2022 - I started to PhD at Univesity of Bakircay Computer Engineering Department.
  • January 2022 - I graduated MSc from Univesity of Ege Computer Engineering Department.
  • November 2019 - I started to work as Computer Vision Researcher at Syntonym.
  • September 2019 - I started to MSc. in Computer Engineering Department at Univesity of Ege.
  • June 2019 - I graduated BSc from Konya Tecnical University as a Computer Engineer.
  • Januery 2014 - I start to study Computer Engineering at Konya Tecnical University.