Open to opportunities

Yasin Esfandiari

@yasinesfandiari

Message

M.Sc. student specializing in Visual Computing and AI research.

Germany

Message

What I'm looking for

I am looking for opportunities that promote innovation, collaboration, and growth in AI and machine learning.

As a dedicated M.Sc. student in Visual Computing at Universität des Saarlandes, I am passionate about advancing the field of artificial intelligence through innovative research. My thesis focuses on the Image Quality-Likelihood trade-off in Diffusion Models, where I explore the balance between image quality and likelihood, aiming to implement new ideas that enhance both aspects.

During my time as a Student Research Assistant at CISPA, I contributed to projects involving Secure Code Generation and Deepfake Text Detection, fine-tuning large language models for practical applications. My experience spans various domains, including Disentangled Representation Learning and Federated Learning, where I analyzed unsupervised methods and metrics, showcasing my ability to tackle complex challenges in AI and machine learning.

With a strong foundation in programming languages such as Python and C++, alongside tools like PyTorch and Keras, I am equipped to develop cutting-edge solutions in computer vision and generative AI. I am eager to leverage my skills and knowledge in a dynamic environment that fosters innovation and collaboration.

Experience

Work history, roles, and key accomplishments

Current

Master Thesis Student

Current

Helmholtz AI

Dec 2023 - Present (1 year 5 months)

Currently working on a thesis project focused on the Image Quality-Likelihood trade-off in Diffusion Models, implementing new ideas to enhance both image quality and likelihood.

Diffusion Models Neural ODEs FID Consistency Models

Student Research Assistant

CISPA – Helmholtz Center for Information Security

Apr 2023 - Sep 2023 (5 months)

Assisted in projects related to Secure Code Generation via LLMs and Deepfake Text Detection, focusing on fine-tuning LLMs for code generation and implementing related research papers.

LLM CodeGen Deepfake Generative AI

Student Research Assistant

CISPA – Helmholtz Center for Information Security

Mar 2022 - Mar 2023 (1 year)

Conducted research on Disentangled Representation Learning and Federated Learning, analyzing unsupervised methods and metrics in data heterogeneity.

Federated Learning Optimization Generative AI Beta TCVAE

Teaching Assistant

Shahid Rajaee Teacher Training University

Jan 2019 - Aug 2019 (7 months)

Supported courses in Advanced Programming with C++, Computational Intelligence, and Data Structures, assisting students in understanding complex concepts.

C++Data Structures Teaching Curriculum Development

Software Developer Intern

Rahkar Hooshmand Iranian

Jun 2018 - Sep 2018 (3 months)

Developed a blog CMS, forum, and eCommerce website using PHP and Laravel, enhancing web development skills and understanding of backend technologies.

PHP Laravel Web Development CMS Ecommerce

Full-stack Web Application Developer

Rahkar Hooshmand Iranian

Jul 2016 - Jan 2018 (1 year 6 months)

Created websites connecting employers with job seekers and providing academic supervision, utilizing full-stack development skills.

Full Stack Developer Web Applications PHP MySQL Laravel

Education

Degrees, certifications, and relevant coursework

Universität des Saarlandes

Master of Science, Visual Computing

2021 -

Grade: 1.6

Activities and societies: Thesis Topic: Image Quality-Likelihood trade-off in Diffusion Models.

Pursuing a Master of Science in Visual Computing with a focus on image quality and diffusion models. Courses include Advanced Image Analysis, High-level Computer Vision, and Machine Learning.

Shahid Rajaee Teacher Training University

Bachelor of Science, Software Engineering

2015 - 2019

Grade: 17.11 / 20

Activities and societies: Thesis Topic: Development and Implementation of an Image Content Descriptor System.

Completed a Bachelor of Science in Software Engineering with a strong focus on image processing and deep learning. Developed an image content descriptor system as a thesis project.