Hyunseok Oh

M.S./Ph.D. Student in

Human-Centered Computer Systems Lab

Dept.of Computer Science and Engieering

Seoul National University

Seoul, Republic of Korea

  I am currently a M.S./Ph.D. student advised by Prof. Youngki Lee at the Department of Computer Science and Engineering, Seoul National University, Republic of Korea. Before the M.S./Ph.D. study, I received a B.S. degree from the Department of CSE, Seoul National University, Republic of Korea. I did an undergraduate internship in Samsung SDI. My major research interest lies in Computer Vision, privacy-preserving Deep Learning and theoretical Machine Learning. Specifically, I aim to develop privacy-preserving ML applications against mass surveillance and involuntary privacy leakage.

E-mail Contact : ohsai@snu.ac.kr

Github alias : ohsai

Publications

• Learning Task-agnostic Representation via Toddler-inspired Learning.

        Kwanyoung Park, Junseok Park, Hyunseok Oh, Youngki Lee, Byoung-Tak Zhang. NeurIPS 2020 Workshop on BabyMind. November 2021.

• 유아모사 베이지안 학습.

        박준석, 박관영, 오현석, 이영기, 장병탁. 한국컴퓨터종합학술대회(KCC). 2020.

• Exploring Image Reconstruction Attack in Deep Learning Computation Offloading.

        Hyunseok Oh, and Youngki Lee. The 3rd International Workshop on Deep Learning for Mobile Systems and Applications(EMDL Workshop on MobiSys). June 2019.

• Privacy Attack in Deep Learning Cloud Server and its Countermeasure

          B.S. Thesis

Awards & Educations

• Seoul National University, 2014-2019

          Bachelor of  Computer Science & Engineering,  Cum Laude.

• National Excellence Scholarship. (Science and Engineering)

          4 years. B.S.

• Accepted to Seoul National University and KAIST as a Priority Chosen in Entrance Examination

Teaching

• M1522.000600 Computer Programming

        CSE, Seoul National University, Fall 2019/2020

• M1522.003300 Advanced Mobile Computing

        CSE, Seoul National University, Spring 2019/2020

Projects

• BabyMind: Building Infant-Mimic Neurocognitive AI

        Seoul National University, 2019/2020

• Secure and Efficient Homomorphic Tensor Encryption for DNN Offloading

        Seoul National University, 2019

• Stochastic Gradient Descent and Multi-layer Perceptron Implementation with Haskell

        Samsung SDI, Spring 2018

• API Elastic Load Balancing with GoLang

        Samsung SDI, Spring 2018

• DQN-based Reinforcement Learning on OpenAI Gym

        Seoul National University, 2018

• Phong Shading-based Efficient Ray-tracing

        Seoul National University, 2018

• Spam Filtering on Review Texts with ensemble of CNN, RNN and Naive Bayes

        SNU Creative Integrated Design Lecture project, 2016

For my works in highschool, refer to link.