K-Darpa Ship Detection Challenge

Deep Learning, Challenge

Participate ship detection challenge hosted by K-DARPA (Defense Advanced Research Projects Agency of Republic of Korea) which is affliate of Agency for Defense Development. The goal is make a segmentation mask given satelite ocean image including negative and positive sample. And classify its type into 5 class.

  • Hosted by K-DARPA
  • AI, Vision
  • Date: Jun 2019 - Present.
Challenge Website

Yonsei Univ. RGB+D Dataset Website

Web

연세대학교 전자공학과 디지털 이미지 미디어 랩 (DIML)의 RGB+D Dataset 웹페이지 제작 프로젝트입니다. One-page layout 및 responsive 기술을 적용하였습니다. github pages를 활용했으며 visitor counter, cluster map등을 적용하였습니다.

Go to Website

Rooftop Party Invitation Website

Web

문화기획팀 '상상실현'의 루프탑파티 웹페이지 제작 프로젝트입니다. 원클릭 카카오페이 결제기능으로 편의성을 극대화하는데 촛점을 두었습니다. github pages를 활용해 서버비용을 절감했고 One-page layout 및 responsive 기술을 적용하였습니다.

Go to Website

"ELSA" : Eldery Caring AI Chatbot

Deep Learning

노인의 생활을 도와주는 능동형 인공지능 챗봇을 개발합니다. 식단 인지와 추천등의 식사 말벗 기능, 가족, 요양사 등 Care Giver와의 연결, 가족과 의사소통을 돕는 인공지능 가족뉴스 요약기능등을 제공합니다.

  • Korea Institute of Science and Technology
  • AI, Vision, NLP, Robotics, AWS
  • Date: 1st Jul. - 28th Nov. 2018
Project Details

Train a Quadcopter how to fly

Deep Learning

I designed an agent that can fly a quadcopter, and then trained it using a reinforcement learning algorithm of your choice. Then, applied the techniques I have learnt in this module to find out what works best, but you will also have the freedom to come up with innovative ideas and test them on my own.

  • Udacity
  • Reinforcement Learning
  • Date: 1st Aug. - 28th Nov. 2018
Github

Generate Faces

Deep Learning

Learned to understand and implement the DCGAN model to simulate realistic images. Then, built a pair of Multi-Layer Neural Networks and made them compete against each other in order to generate realistic faces.

  • Udacity
  • GAN, TensorFlow
  • Date: 1st Aug. - 28th Nov. 2018
Github

Generate TV Script

Deep Learning

In this project, I built my own Recurrent Networks and Long Short-Term Memory Networks with PyTorch. I performed sentiment analysis and generate new text, and use recurrent networks to generate new text that resembles a training set of TV scripts.

  • Udacity
  • RNN, LSTM
  • Date: 1st Aug. - 28th Nov. 2018
Project Details

Dog Breed Classifier

Deep Learning

I learned how to build a pipeline to process real-world, user-supplied images. Given an image of a dog, my algorithm will identify an estimate of the canine’s breed. If supplied an image of a human, the code will identify the resembling dog breed.

  • Udacity
  • CNN, Transfer Learning, Autoencoders
  • Date: 1st Aug. - 28th Nov. 2018
Github

Bike Sharing Prediction Project

Deep Learning

I learned neural networks basics, and build my first network with Python. I defined and train a multi-layer neural network, and use it to analyze real data. In this project, I built trained neural networks from scratch to predict the number of bike-share users on a given day.

  • Udacity
  • Multi Layer Perceptron, Keras
  • Date: 1st Aug. - 28th Nov. 2018
Github

Personal Hompage/Blog Project

Web

I built my own portfolio homepage and blog using github-pages and jekyll. I learned how to use HTML, CSS and javascripts. And I also get used to manage my projects and code with github.

  • Personal Project
  • HTML, CSS, Javascript, Jekyll, Github
  • Date: 1st June, 2018
This Blog!

Withblock Website Outsourcing

Web

I had requested to make blockchin company website with github and jekyll. This project is my first paid outsourcing one with my web development skill.

  • Outsourcing
  • HTML, CSS, Javascript, Jekyll, Github Pages
  • Date: 1st July, 2018
Go to Demo site