image

Lin Magari

Graduate Student

About Me

Work Experience

Apr 2024 - Sep 2024
Tokyo, Japan

iOS Software Engineer Intern, Apple

  • SpringBoard
Aug 2022 - Dec 2022
Tokyo, Japan

Software Engineer Intern, Google

  • Worked on the ChromeOS networking team.
Apr 2022 - Sep 2022
Iizuka, Fukuoka, Japan

Teaching Assistant, Kyushu Institute of Technology

  • Assisted students on their experiment assignments with basic knowledge of Python and machine learning.
March 2022
Iizuka, Fukuoka, Japan

Research Assistant, Kyushu Institute of Technology

  • Conducted experiments related to segment clouds from satellite images.
Sep 2021
Iizuka, Fukuoka, Japan

Research Assistant, Kyushu Institute of Technology

  • Conducted experiments for image segmentation tasks.

Education

Apr 2022 - Sep 2024
Iizuka, Fukuoka, Japan

MSc in Computer Science, Kyushu Institute of Technology

  • Graduate School of Computer Science and Systems Engineering
  • Research Theme:
    • Heavy snow cloud segmentation in meteorological satellite images based on semi-supervised image segmentation tasks.
  • Conference Presentations
    • Magari, Lin, Terumasa Tokunaga, and Kazue Suzuki. "Heavy Snow Cloud Detection in Satellite Images Based on Semi-Supervised Image Segmentation." (2023).
May 2022 - Jun 2022
Remote

Google STEP Development Course

  • An 8-week training program to learn the basics of computer science. The course covers the basics of data structures, algorithms, graph theory, theory of computation, and system architecture.
Apr 2018 - March 2022
Iizuka, Fukuoka, Japan

BSc in Computer Science, Kyushu Institute of Technology

  • School of Computer Science and Systems Engineering
  • Department of Artificial Intelligence, Media Informatics Course
  • Undergraduate Thesis:
    Heavy Snow Cloud Segmentation in NOAA Satellite Images Based on Semi-Supervised Binary Classification.
  • enPiT AiBiC: Future Prediction (Sep 2020)
    • Developed dissatisfaction prediction system using NLP techniques.
    • Trained language model using Python, MeCab, and Pandas on complaints data from fumankaitori.com in travel category.
    • Implemented frontend using HTML, CSS, and Flask framework.
    • Outputs answers based on cosine similarity ranking, providing insights into potential dissatisfaction factors.
    • Achieved 1st place in enPiT 2020 competition.

Projects

Heavy snow cloud detection in satellite images based on semi-supervised learning

Detect the heavy snow cloud from satellite images obtained by meteorological satellites, based on semi-supervised image segmentation to estimate the amount of snowfall in Antarctica, the Arctic, and the airspace surrounding Japan. (Sep 2021 - Present)

Discord Bot

Created a discord bot to understand the basic discord APIs. The name of the bot is Cassie's App. It can say Hello to others, as well as add reactions to messages.
Periods of Development: 2 weeks.

Tech Stack

  • Python
  • Java
  • C++
  • HTML / CSS
  • JavaScript
  • R
  • Prolog
  • Image Segmentation
  • Keras / Tensorflow
  • Chromium (Mojo)
  • Git
  • Discord API (with Python)

Recent Learning

  • Objective-C
  • Swift
  • SwiftUI
  • UIKit
  • SQL / MySQL / PostgreSQL
  • Pandas
  • PyTorch
  • Kotlin (TODO)

Languages

  • Japanese
  • Chinese
  • English

Volunteer

Waffle.org

Volunteer as a bootcamp TA to support a team for middle and high school students in Technovation Girls Japan 2023. (Jan 2023)

Interests

Software Engineering

Backend Development

Platform Development

Full-stack Development

Mobile App Development

Image Recognition

Image Segmentation