Hyunseok Oh

Hyunseok Oh

Personal Profile

Ph.D. candidate in the Department of Computer Science and Engineering at Seoul National University with a research focus on human-centered AI, biologically inspired AI, next-generation AI technologies like Spiking Neural Network, resource-efficient AI, and theoretical machine learning. Specializes in developing novel methodologies and efficient AI systems, contributing to both theoretical advancements and practical implementations.

Education

Seoul National University
Seoul, Korea
M.S./Ph.D. in Computer Science and Engineering
Mar. 2019 -
Advisor: Youngki Lee
Seoul National University
Seoul, Korea
B.S. in Computer Science & Engineering
Mar. 2014 - Feb. 2019

Publications

  1. [C8] VoiceSeeker: Energy-Efficient and Accurate Audio-Visual Target Speaker Separation on Mobile Devices
    Juheon Yi, Kyungjin Lee, Hyunseok Oh and Youngki Lee
    IEEE Pervasive Computing 2025 (PerCom'25).
  2. [C7] Sign Gradient Descent-based Neuronal Dynamics: ANN-to-SNN Conversion Beyond ReLU Network [PDF] [Slide] [Code]
    Hyunseok Oh and Youngki Lee
    International Conference on Machine Learning 2024 (ICML'24).
  3. [C6] Papez: Resource-Efficient Speech Separation with Auditory Working Memory [PDF] [Code]
    Hyunseok Oh, Juheon Yi, and Youngki Lee
    IEEE International Conference on Acoustics, Speech and Signal Processing 2023 (ICASSP'23).
  4. [C5] VECA: A New Benchmark for General Cognitive Development. [PDF] [Code]
    Kwanyoung Park*, Hyunseok Oh*, and Youngki Lee (*Co-primary)
    AAAI Conference on Artificial Intelligence 2022 (AAAI'22) [Oral Presentation, top 4.15%].
  5. [C4] On the Importance of Critical Period in Multi-stage Reinforcement Learning. [PDF]
    Junseok Park, Inwoo Hwang, Min Whoo Lee, Hyunseok Oh, Minsu Lee, Youngki Lee, Byoung-Tak Zhang
    ICML Complex Feedback in Online Learning Workshop 2022 (ICMLW'22).
  6. [C3] Toddler-Guidance Learning: Impacts of Critical Period on Multimodal AI Agents [PDF]
    Junseok Park, Kwanyoung Park, Hyunseok Oh, Ganghun Lee, Minsu Lee, Youngki Lee, Byoung-Tak Zhang
    ACM International Conference on Multimodal Interaction 2021 (ICMI'21) [Acceptance Rate: 13%].
  7. [C2] Learning Task-agnostic Representation via Toddler-inspired Learning [PDF]
    Kwanyoung Park, Junseok Park, Hyunseok Oh, Byoung-Tak Zhang, and Youngki Lee
    NeurIPS Workshop on BabyMind (NeurIPSW 2020).
  8. [C1] Exploring Image Reconstruction Attack in Deep Learning Computation Offloading [PDF]
    Hyunseok Oh, Youngki Lee
    ACM International Workshop on Deep Learning for Mobile Systems and Applications (MobiSysW'19).

Awards \& Honors

BrainKorea21 Star Student Researcher 2023, Seoul National University
2023
Graduate Student Instructor Scholarship, Seoul National University
2019 - 2021
National Academic Excellence Scholarship
2014 - 2018
Science and Engineering

Skills

Work Experience

Microsoft Research
Redmond, Washington, United States
Research Intern (Advisor: Matthai Philipose)
June. 2025 - Sep. 2025
  • Formal Document Processing.
Microsoft Research
Redmond, Washington, United States
Research Intern (Advisor: Matthai Philipose)
June. 2024 - Sep. 2024
  • Logical-Z3: LLM-based Natural Language Theory Solver for Z3 theorem prover.
Samsung Electronics, Mobile Division
Suwon, Korea
Student Intern (Advisor: Euteum Park)
June. 2018 - Sep. 2018
  • Developed a prototype Golang-based end-to-end API Gateway for Samsung Cloud.

Academic Services

NeurIPS Conference Main Track Reviewer
2025
ICML Conference Main Track Reviewer
2025
ICLR Conference Main Track Program Committee
2025,2026
AAAI Conference Main Track Technical Program Committee
2023,2024,2025,2026
ICASSP Conference Main Track Technical Program Committee
2023,2024,2025

Teaching Experience

Teaching Assistant, Seoul National University
Fall, 2019-2020
M1522.000600 Computer Programming
Teaching Assistant, Seoul National University
Spring, 2019
M1522.002600 Advanced Mobile Computing

Mentored Students

Jimin Chae (SNU Undergraduate, Winter )
2022-2024
Spiking Neural Network
Kwanyoung Park (SNU Undergraduate)
2019-2023
Toddler-inspired Artificial Intelligence [C2,C3,C5]
Younghoon Choi (UC Berkeley Undergraduate)
2023
Listener Identification in Parent-child Interaction

Invited Talks

Chosun Center for Data Science in Humanities Annual Workshop
2024
Chosun University, Gwangju, Korea
CSE AI Summer School
2023
Seoul National University, Seoul, Korea
AIIS Spring Retreat
2022
Seoul National University, Seoul, Korea

Other Projects

Spiking Fourier Neural Operator for Energy-Efficient PDE Solving
First fully spiking neural network architecture designed for modeling infinite-dimensional operators.
Nov. 2024 -
Hyunseok Oh, Youngki Lee
Automated Classification of Child-directed Speech in Parent-Child Interaction
Listener identification with speech classification DNN.
Apr. 2023 -
Hyunseok Oh, Eon-Suk Ko, Youngki Lee
Implicit Facial Attribute De-identification
GAN-based face image editing technique that protects all the sensitive attributes implicitly
Mar. 2020 - Feb. 2023
Hyunseok Oh, Youngki Lee
SilverStone: Privacy-Preserving and Efficient DNN Computation Offloading
Cloud-offloaded encrypted DNN inference with homomorphic tensor encryption scheme
Jan. 2019 – Dec. 2019
Hyunseok Oh, Eunsu Baek, Youngki Lee

References

Youngki Lee, Associate Professor, Seoul National University
youngki.lee@gmail.com
Matthai Philipose, Senior Principal Researcher, Microsoft Research
matthaip@microsoft.com