Dayeon Ku, an integrated MS-PhD student in the lab, is featured in the student interview section of the 2027 Graduate Lab Guide published by the School of Electrical Engineering and Computer Science at GIST. She talks about why she came to GIST, what she works on, and what the lab is like.
Q. What made you decide to pursue graduate study at GIST?
It started from the thought that text alone carries only a fraction of what people actually exchange. I studied data science as an undergraduate and worked on natural language processing, then my interest moved to speech and audio, and I decided to go to graduate school to learn it properly. Seeing the breadth of Prof. Hong Kook Kim's work on audio AI, including accessibility, I felt it matched the direction I wanted. Having lived abroad for a long time, coursework in Korean would have been a burden, so the fact that GIST teaches in English weighed heavily too.
Q. Could you briefly introduce your research interests?
I work on automatic dubbing — carrying the lines in a video naturally into another language. So far I have worked on making translation account for a speaker's expression and tone and the mood of the scene, and on locating pauses between utterances accurately even in noisy conditions. I am currently working on multimodal lip-sync.
Q. What are the strengths and weaknesses of graduate study at GIST, as a current student?
The biggest is being able to focus entirely on research without worrying about tuition or living costs. I received a best paper award at a domestic conference six months in, and had a paper accepted at an international venue within a year and a half; that pace is possible because of this environment. The downside is how strict security is. Without an access card it is hard to move around the buildings, and the network is tightly controlled, so things occasionally get blocked. Put another way, it means the research environment is well managed.
Q. What is the atmosphere in the lab like?
There is no hierarchy to navigate and no need to read the room. Everyone has a different topic, so we mostly focus on our own work, and nothing gets done because someone was told to do it. The professor is easy to talk to but still points out details from a researcher's perspective, which helps a lot when setting a direction or when I am stuck. He often buys meals or coffee, so there are plenty of chances to talk naturally. Being able to talk about the research I want to do and push it forward is the biggest strength.
View the original PDF