cv

Basics

Name Parham Saremi
Label MSc Student at Mila | McGill University
Email parham.saremi@mail.mcgill.ca
Summary Graduate student at Mila and McGill University supervised by Prof. Tal Arbel, with research interests in computer vision, generative modeling, and medical imaging.

Education

  • 2024.05 - Present

    Montreal, Canada

    Master of Science
    McGill University | Mila
    Electrical and Computer Engineering
  • 2018.09 - 2023.06

    Tehran, Iran

    Bachelor of Science
    Sharif University of Technology
    Computer Engineering

Work

  • 2024.05 - Present

    Montreal, Canada

    Graduate Research Student
    Mila - Quebec Artificial Intelligence Institute
    Conducting research in medical imaging and generative modeling under the supervision of Prof. Tal Arbel.
    • Working on Diffusion Models for explainability and uncertainty quantification in medical images.
    • Trained stable-diffusion on medical data and fine-tuned using policy optimization to improve alignment with prompts by 11%
    • Worked on Distillation of Diffusion Models for faster inference and deployment.
  • 2022.07 - 2023.09
    .NET Developer (Part-time)
    Node Effect
    Worked as a .Net developer and contributed to C# and F# open-source projects.
    • Developed a MAUI-based application (C#) and contributed to the MAUI Linux project.
    • worked on Bitcoin Lightning-related projects.
  • 2022.07 - 2023.07
    Research Intern
    ETH Zürich – AIT Lab
    Worked as a research intern on Human-Object Interaction and 3D reconstruction.
    • Designed and implemented pipelines to generate 3D human and object models from single images while reasoning about their interaction.
    • Used the decoder of a pre-trained VQ-VAE model to generate object meshes from images.
  • 2021.12 - 2022.07
    Research Intern
    EPFL – VITA Lab
    Worekd as a researh intern on Human Pose Forecasting.
    • Developed and maintained UnPOSed, an open-source toolbox for forecasting a sequence of human pose in future.
    • Designed and evaluated a method for human motion prediction that improved results of various SOTA models by 2-5% across multiple datasets.
  • 2021.11 - 2022.07
    Research Collaborator – B.Sc. Project
    University of California Irvine & Sharif University of Technology
    Joint reserach project between UCI and Sharif University of Technology for my B.Sc. thesis.
    • Worked with vision-language models, Transforemrs, and Graph Neural Networks.
    • Designed and developed a multimodal model using GNNs and Transformers to predict cuisine using ingredient information and recipes.
  • 2021.07 - 2021.10
    Research Intern
    Imperial College London
    Investigated and experimented with new methods for the verification of Neural Networks.

Skills

Machine Learning
Python
PyTorch
PyTorch3D
Numpy
Pandas
Scikit-learn
Matplotlib
HuggingFace
Other
Linux
Git
GitHub
C/C++
Java
C#
F#
LaTeX
Excel

Languages

English
Fluent (TOEFL 111)
Persian
Native