Hallym University / School of Software

Information Theory and
Statistical Inference Lab

We study the mathematical foundations of communication, learning, and inference. Our work asks how much information is needed—and how it should be represented—to communicate reliably, learn from data, and make decisions under uncertainty.

Join us

Contact Sung Hoon Lim

Graduate students and postdoctoral researchers

We are recruiting graduate students and postdoctoral researchers interested in information theory, communication, compression and generative learning, or statistical inference and sensing.

Please email shlim@hallym.ac.kr with your CV and a brief description of your research interests. Postdoctoral applicants are also welcome to include links to representative publications. Research fit, available support, and start dates will be discussed individually.

대학원생 및 박사후연구원 모집

정보이론, 통신, 압축 및 생성모델, 통계적 추론과 센싱 분야에 관심 있는 대학원생과 박사후연구원을 모집합니다. 이력서와 관심 연구 분야를 간단히 정리하여 shlim@hallym.ac.kr로 보내 주세요. 연구 주제, 지원 조건 및 합류 시기는 개별 상담을 통해 안내합니다.

Research

  1. 01

    Information theory and communication

    Coding strategies and fundamental limits for multiuser networks, joint source–channel communication, and distributed information processing.

  2. 02

    Compression and generative learning

    Rate–distortion–perception tradeoffs, diffusion-based compression, and semantic representations that preserve the information relevant to a task. Generative watermarking studies reliable information embedding through pretrained generators, with distribution preservation, secret-key constraints, and robustness to regeneration.

  3. 03

    Statistical inference and sensing

    Quickest change detection and integrated sensing and communication: understanding the tradeoffs between data rate, detection delay, and error probabilities.

People

Director

Sung Hoon Lim 임성훈

Associate Professor, School of Software
Hallym University · IEEE Senior Member

EmailORCIDGoogle Scholar

Students

Recent publications

All publications

Recent preprint / September 2026

On the Information-Theoretic Limits of Latent-Space Watermarking Through Pretrained Generators

Jinwan Jeon, Minju Lee, and Sung Hoon Lim

Rate–key limits and regeneration robustness for watermarking through a fixed pretrained generator.

Recent journal articles. Journal archive / Conference papers

Research funding

Current project

Diffusion-based image compression

정보이론적 지각성을 반영한 디퓨전 네트워크 기반 영상 압축 코덱 원천기술 연구

National Research Foundation of Korea (NRF)
2026-03-01 – 2029-02-28

Past projects (5)

Multiuser compression and federated learning

VQ-VAE 기반 다중사용자 정보압축 부호체계 연구 및 연합학습 적용

National Research Foundation of Korea (NRF)
2023-09-01 – 2026-02-28

Semantic extraction and reconstruction

정보이론 활용 시맨틱 추출 복원 기계학습 코드 개발

Electronics and Telecommunications Research Institute (ETRI)
2023-10-30 – 2024-02-29

Distributed edge intelligence

분산 엣지 인공지능 네트워크를 위한 정보압축 및 정보전송 기계학습 알고리즘 원천기술 연구

National Research Foundation of Korea (NRF)
2020-06-01 – 2023-02-28

Information computation and communication

정보 연산 및 정보 전달에 대한 근본적 이론 연구

Hallym University
2019-10-01 – 2020-09-30

Distributed computing for maritime information systems

분산 컴퓨팅 및 정보전송을 위한 근본적 이론연구 : 해양 정보 시스템 적용

National Research Foundation of Korea (NRF)
2019-03-01 – 2020-02-29

Contact

shlim@hallym.ac.kr

Sung Hoon Lim, Lab Director

+82-33-248-2358
Faculty office: Engineering Building, Room 1242

School of Software
Hallym University

1 Hallymdaehak-gil, Chuncheon
Gangwon-do 24252, South Korea

강원특별자치도 춘천시 한림대학길 1

Campus map