# About

<figure><img src="https://704680750-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-M01K_pXDa9D-2q3E1yB%2Fuploads%2F8Ovvz67n3PhOhFPOGOX4%2FInsuJeon2.jpg?alt=media&#x26;token=2c8658c8-6717-4c81-96be-a536f3a55f1c" alt="" width="375"><figcaption></figcaption></figure>

## Insu Jeon

Hello, I am a person who is interested in innovating the world with Artificial Intelligence (AI) technology for human well-being and happiness.

I graduated from [UCLA](https://insujeon.gitbook.io/me/education/ucla) with a major in [Computer Science](https://samueli.ucla.edu/) and a minor in [Statistics](https://stats.oarc.ucla.edu/).

I graduated from [Seoul National University](https://insujeon.gitbook.io/me/education/snu) with a Ph.D. in [Computer Science and Engineering](https://cse.snu.ac.kr/en) under the supervision of [Prof. Gunhee Kim](https://vision.snu.ac.kr/gunhee/index.html#top) in the [Vision & Learning Lab](https://insujeon.gitbook.io/me/work-experiences/vision-and-learning-lab-snu.). I was under [Prof. SI Yoo](https://cse.snu.ac.kr/professor/%EC%9C%A0%EC%84%9D%EC%9D%B8) in the Artificial Intelligence Laboratory during my master's period.

I have participated in various Artificial Intelligence (AI) projects, including [Computer Vision](https://en.wikipedia.org/wiki/Computer_vision), [Natural Language Processing](https://en.wikipedia.org/wiki/Natural_language_processing), [Bayesian deep learning](https://en.wikipedia.org/wiki/Bayesian_network), [Generative model](https://en.wikipedia.org/wiki/Generative_model), [Meta-Learning](https://en.wikipedia.org/wiki/Meta-learning_\(computer_science\)), [Federated Learning](https://en.wikipedia.org/wiki/Federated_learning), and [Large Language Model](https://en.wikipedia.org/wiki/Large_language_model).

I have an [entrepreneurship ](https://insujeon.gitbook.io/me/work-experiences/rippleai)experience where I learned valuable lessons to be a tech leader.

Here is my [CV](https://insujeon.github.io/data/InsuJeon_cv2023.pdf).

{% hint style="info" %}
If you have a question about me, ask to Chat-GPT using the search box of this web page (press ctrl + k and select the Lens option) or email me.
{% endhint %}

## Contact

:e-mail: <insuj3on@gmail.com>

## Education

**Ph.D.:** Department of Computer Science and Engineering, [**Seoul National University**](https://insujeon.gitbook.io/me/education/snu)**,** 2013 - 2023

**B.S.:** Major in Computer Science, Minor in Statistics, [**University of California in Los Angeles**](https://insujeon.gitbook.io/me/education/ucla)**,** 2009 - 2012

## Work Experiences

**Lead, AI Tech Lab & Product,** [**QRAFT**](https://qraftec.com/)**. Oct** 2023 –  present

* Developing a Large Language Model (LLM)-based chatbot application for financial services: LLM-based summary of overseas disclosures and financial counseling chatbot for customers.

**AI Researcher,** [**Vision & Learning Laboratory**](https://insujeon.gitbook.io/me/work-experiences/vision-and-learning-lab-snu.)**, SNU.** Mar 2017 – Sep 2023

* Conducted advanced research in Generative models, Natural Language Processing (NLP), and Bayesian meta-learning.

**AI Researcher,** [**Everdoubling**](https://insujeon.gitbook.io/me/work-experiences/everdoubling)**.** Jun 2021 – Dec 2021

* Participated in AI Grand Challenge; developed a math problem-solving AI engine using a General Language Model.

**Chief Technology Officer (CTO),** [**RippleAI**](https://insujeon.gitbook.io/me/work-experiences/rippleai)**.** Feb 2018 – Dec 2019

* Managed a team of 9 developers and engineers; developed an Instagram comment-generating bot.

**Machine Learning Researcher, Artificial Intelligence Laboratory, SNU.** Sep 2012 – Sep 2016

* Developed ML algorithms for computer vision tasks such as defect detection, super-resolution, and registration.

## Published Paper

Minui Hong, Junhyeog Yun, **Insu Jeon,** Gunhee Ki&#x6D;**.** "FedAutoAug: Augment Local Data via Shared Policy in Federated Learning." *CVPR*, 2024. (under review)

**Insu Jeon**, Minui Hong, Junhyeog Yun, Gunhee Ki&#x6D;**.** "Federated Learning via Meta-Variational Dropout." *NeurIPS*, 2023. \[[paper](https://openreview.net/forum?id=VNyKBipt91)]\[[code](https://github.com/insujeon/MetaVD)]

**Insu Jeon**, Junhyeog Yun, Minui Hong, and Gunhee Kim. "Data Augmentation via Generation Model in Military Aircraft Classification." *KIMST*, 2023.

**Insu Jeon**, Youngjin Park and Gunhee Kim. "Neural Variational Dropout Processes." *ICLR*, 2022. \[[paper](https://openreview.net/forum?id=lyLVzukXi08)] \[[code](https://github.com/insujeon/NVDPs)] \[[project](https://insujeon.gitbook.io/me/published-papers/neural-variational-dropout-processes)]

**Insu Jeon**, Wonkwang Lee, Myeongjang Pyeon and Gunhee Kim. "IB-GAN: Disentangled Representation Learning with Information Bottleneck Generative Adversarial Networks." *AAAI*, 2021. \[[paper](https://ojs.aaai.org/index.php/AAAI/article/view/16967)] \[[code](https://github.com/insujeon/IB-GAN)] \[[project](https://insujeon.gitbook.io/me/published-papers/ib-gan)]

**Insu Jeon**, D Kang, SI Yoo. "Blind image deconvolution using Student's-t prior with overlapping group sparsity." *ICASSP*, 2017. \[[paper](https://ieeexplore.ieee.org/document/7952470)] \[[pdf](https://arxiv.org/abs/2006.14780)] \[[project](https://insujeon.gitbook.io/me/published-papers/blind-image-deconvolution-using-students-t-prior-with-overlapping-group-sparsity)]

**Insu Jeon**, SI Yoo. "Spatial kernel bandwidth estimation in background modeling." *ICMV*, 2016. \[[paper](https://www.spiedigitallibrary.org/conference-proceedings-of-spie/10341/1/Spatial-kernel-bandwidth-estimation-in-background-modeling/10.1117/12.2268512.short?SSO=1)]

## Projects

**Unsupervised Learning-based Data Generation Research, Agency for Defense Development (ADD).** Jun 2022 – Present

* Improved military object recognition performance by 10% via Generative model-based data augmentation.

**Neural Processing System Research, Samsung Advanced Institute of Technology.** Mar 2018 – Sep 2019

* Contributed to Samsung’s core AI vision technology and organized group activities for researchers.

**Computer Vision Projects, Samsung Device Solutions Institute.** Mar 2013 – Sep 2017

* Optimized defect-monitoring systems in semiconductor display (SEM/OLED) production lines.

## Awards

**1th Kbig-contest – National Information Society Agency.** Dec 2013

* Developed Twitter hot issue forecaster using NLP algorithm and placed an encouragement award.

## Teaching Experience

**Special Lectures on Bayesian Data Analysis and Statistical Inference – GSSHOP.** May 2019

* Delivered lectures on Bayesian theory and statistical inference techniques for commercial data analysis. \[[part1](https://insujeon.github.io/data/BayesForML_Day1.pdf)] \[[part2](https://insujeon.github.io/data/BayesForML_Day2.pdf)] \[[part3](https://insujeon.github.io/data/BayesForML_Day3.pdf)]

**Practical Guide to Deep Learning – Korea Banking Institute.** Mar 2019

* Conducted lectures on Deep Learning and Natural Language Processing.

**Introduction to Generative Model with PyTorch – Fastcampus.** Sep 2017 – Sep 2018

* Taught courses on Deep Generative model and Bayesian Deep Learning. \[[code](https://github.com/insujeon/Hello-Generative-Model)]

**Special Issues in Machine Learning and Deep Learning – Seokyong University.** Jun 2017

* Performed lectures on Modern developments in Machine Learning, Deep Learning, and Artificial Intelligence.

**Prerequisite Courses for Artificial Intelligence – SNU 4th Industrial Revolution Academy.** May 2017

* Prerequisite courses for understanding Artificial Intelligence - Linear Algebra, Probability, and Statistics. \[[part1](https://insujeon.github.io/data/LinearAlg.pdf)]

## Technical Skills

**Computer Skills**

* Python, C/C++, Java, JavaScript, Objective C, OpenMP, CUDA, HTML, Bash, Windows, Mac OS, Linux, Huggingface, Langchain

**Languages**

* Korean (native), English (proficient).