AI Stochastic Process 3 Korean course Major 본원 -

Course Name

AI Stochastic Process

Summary

Lecture Objectives

The aim of the lecture is to understand the theory of stochastic processes adequate to modern artificial intelligence such as the generated model theory. In this lecture, the main contents are the modern stochastic process theory, stochastic differential theory, and machine learning from the perspective of stochastic analysis. The detailed contents of the lecture are as follows: - From the fundamental to the advanced theory of stochastic/statistical analysis for machine learning - Machine learning and artificial intelligence theory from the viewpoint of nonlinear filtering theory - Stochastic differential equation as the advanced theory of stochastic process and modern filtering theory based on stochastic analysis. - The data processing methodology in artificial intelligence and computer science based on a stochastic perspective.

Introduction to Human-Computer Interaction 3 Korean course Major 본원 -

Course Name

Introduction to Human-Computer Interaction

Summary

Lecture Objectives

In this course, we study fundamental concepts of HCI and considerations for good HCI design. We also have a look on HCI implemented via various modalities in a technical view: such as gestures, gaze, and haptic as well as visual interfaces on touchscreens. The evaluation will be conducted with one report assignment and one presentation of paper review. Some knowledge on programming can be helpful to better understanding of the content of the class, but it is not necessary for the assignment.

Advanced Deep Learning 3 Korean course Major 본원 -

Course Name

Advanced Deep Learning

Summary

Lecture Objectives

This course aims to explore the latest research trends in deep learning. Deep learning is one of the most actively researched fields recently. This course is designed so that students who have taken the basics in the Deep Learning course can cover more advanced topics in depth and use them for research. Specifically, we will review the recent research trends and latest research papers in deep learning, and if necessary, deliver related theories and background knowledge.

Deep Learning 3 Korean course Major 본원 -

Course Name

Deep Learning

Summary

Lecture Objectives

This class aims at introducing the world of deep learning. Through this class, the students are expected to understand the basics and the cores of neural networks including error back propagation and deep learning. Topics will include artificial neural networks basics, convolutional neural networks, generative models( autoencoders, generative adversarial networks), recurrent neural networks including lstm and possibly recent advances in this field, if time allows.As a final note, it course is recommended to the students who are 1) interested in deep learning, of course 2) has a basic knowledge on Python or be willing to learn it quickly.

Deep Reinforcement Learning 3 Korean course Major 본원 -

Course Name

Deep Reinforcement Learning

Summary

Lecture Objectives

This lecture will cover both 1) Basics of RL 2) Advanced RL ＆ recent developments. The course will emphasize the development of intuition relating the mathematical theory of reinforcement learning to the design of human-level artificial intelligence.Students should have quite a good understanding of Python, Pytorch, and Deep learning. If you don’t have taken the “Deep learning” class or don’t understand deep learning, I don’t recommend you to take this lecture.

Advanced Video Understanding 3 Korean course Major 본원 -

Course Name

Advanced Video Understanding

Summary

Lecture Objectives

This lecture deals with computer vision techniques for video understanding as a process of advanced computer vision. Specifically, we address techniques for understanding videos based on specific classes of human actions, such as action recognition on well-trimmed videos and action localization and detection on untrimmed videos, techniques for understanding videos based on queries described by natural sentences, such as temporal moment localization and video QA, and techniques for video and language representation learning. This course includes theoretical lectures as well as practical exercises that require programming. Through this, the purpose of this study is to enable students to learn temporal modeling methods for understanding videos semantically.

Understanding and Applications of Image Processing 3 Korean course Major 본원 -

Course Name

Understanding and Applications of Image Processing

Summary

Lecture Objectives

The aim of this course is to learn about image processing theories and their applied technologies. In this course, students learn about image processing technology through visual results. Based on the learner's background knowledge, there may be a simple programming practice for image processing applications.

Speech Signal Processing 3 Korean course Major 본원 -

Course Name

Speech Signal Processing

Summary

Lecture Objectives

This class aims at introducing the various techniques on speech signal processing and automatic speech recognition(ASR). Speech signal processing is a classical field of research which deals digital signal processing as well as speech enhancement, array signal processing and blind signal separation. Based on the theories of speech signal processing, various technologies of ASR will be explored ranging from the traditional Hidden Markov Model to the most recent end-to-end ASR method. As an excercise the attendees will build Korean end-to-end ASR system using an open-source toolkit and open corpus.

Embedded Deep Learning 3 Korean course Major 본원 -

Course Name

Embedded Deep Learning

Summary

Lecture Objectives

In order to perform deep learning in an embedded system, we will understand the characteristics of the embedded system, and look at parallel processing techniques to accelerate deep learning in the embedded system. In addition, we analyzes the constraints of the embedded system and learns an embedded-based lightweighting technique to lighten the existing deep learning model.

Optimized Learning Theory 3 Korean course Major 본원 -

Course Name

Optimized Learning Theory

Summary

Lecture Objectives

The aim of the lecture is to understand the knowledge of the mathematics correspoding nonlinear optimization, various induction technique of nonlinear optimization algorithms, and the solution through the machine learning algorithm based on a non-linear optimization for enginnering problems.issues based on nonlinear optimization.

Computer Vision 3 Korean course Major 본원 -

Course Name

Computer Vision

Summary

Lecture Objectives

This lecture is an introductory and in-depth course on computer vision, and deals with image geometry and image understanding. Specifically, it aims to provide an understanding of image formation and imaging processes, color and color models, image invariant features and their matching, pattern recognition, motion estimation and object tracking, geometric modeling of images, and geometric interpretation methods of images.

Mathematical Tools for Computer Vision 3 Korean course Major 본원 -

Course Name

Mathematical Tools for Computer Vision

Summary

Lecture Objectives

This course introduces essential mathematical tools used in scientific research such as mathematical basics(PCA, SVD, Taylor expansion, Gradient, Hessian, etc.), geometric operations with vectors, linear systems and matrix operations, parameter estimation techniques, mathematical optimization techniques, and machine learning applications. The ultimate purpose of this course is to enable the students to use these mathematical tools in their own research. A programming assignment is given for each class and basic programming skills(C++ or Python) are required to perform the assignments.

Pattern Recognition and Machine Learning 3 Korean course Major 본원 -

Course Name

Pattern Recognition and Machine Learning

Summary

Lecture Objectives

The aim of this course is to learn about pattern recognition theories and their applied technologies. This course includes conceptual explanations and mathematical expressions for pattern recognition. Based on the learner's background knowledge, there may be a simple programming practice for pattern recognition application.