1. Research Area
- Visual & Biomedical Computing
Visual computing is a field that includes computer vision, computer graphics, HCI, VR/AR and machine learning. Biomedical computing contains a study on healthcare using various biometric Information (EEG, posture, gait, sleep information, heart rate, electromyogram), analyzing medical images and biological data, such as MRI and DNA/protein sequence.
2. Research Overview
We are studying to understand data(visual and bimedical data). Especially, solving problems such as facial emotion recognition, machine vision, behavior analysis and understanding x-objects in videos, and visual speech recognition through deep learning.
√ Analysis of camera images for autonomous driving
It is a study of how to recognize roads, other cars, and pedestrians and determine the speed or direction of movement for the autonomous driving of cars. we are conducting a study using a model car.
√ Reinforcement Learning
It is a study on artificial intelligence that selects the most rewarding behavior in the present state in a environment such as game. Our goal is to propose a new learning model so that we can reach high scores with less training.
√ Few-Shot Learning
When classifying images through deep learning, computers perform better than humans. However, deep learning requires a lot of data and takes a long time to learn. Few-Shot Learning is a way to solve problems with less data. We are studying on applying integrated few-shot learning into deep learning based visual speech recognition.
√ Facial emotion recognition
It is a study to analyze facial expression changes and find out what the current state of emotion is. The main research is to develop analysis methods and deep learning models suitable for human facial characteristics. In addition, we are conducting to understand the relationship between heart rate and emotion
3. Research Achievements
Guang Jin, Yohwan Noh, and DoHoon Lee, “Model-Based Reinforcement Learning with Discriminative Loss,” Journal of KIISE, Vol. 47, No. 6, pp. 547-552, 2020.
Boubenna Hadjer, and DoHoon Lee, "Image-based emotion recognition using evolutionary algorithms," Biologically inspired cognitive architectures, 24, pp. 70-76, 2018