Tae-Hyun Oh (0000-0003-0468-1571)

Personal information

Biography

Tae-Hyun Oh is an associate professor at the School of Computing, KAIST, Korea, since 2025. Before joining KAIST, he was affiliated with Dept. EE, POSTECH, Korea from 2020, Facebook AI Research in 2019, and he was a postdoctoral associate at MIT CSAIL from Aug, 2017. He received the B.E. degree (The highest ranking) in Computer Engineering from Kwang-Woon University, South Korea in 2010, and the M.S. and Ph.D degrees in Electrical Engineering from KAIST, South Korea in 2012 and 2017, respectively. He was a research intern in the Visual Computing group, Microsoft Research, Beijing and in the Cognitive group, Microsoft Research, Redmond in 2014 and 2016, respectively. He was a recipient of Microsoft Research Asia fellowship, a gold prize of Samsung HumanTech thesis award, BMVC 2024 Best Poster Award, Qualcomm Innovation awards, and top research achievement awards from KAIST. His research interests include robust computer vision and machine learning.

Activities

Collapse all

KAIST: Daejeon, KR

2025-02-01 to present (School of Computing)

Employment

Tae-Hyun Oh

https://orcid.org/0000-0003-0468-1571

Personal information

Emails & domains

Verified email addresses

Verified email domains

Websites & social links

Other IDs

Countries

Biography

Activities

Employment (5)

KAIST: Daejeon, KR

Pohang University of Science and Technology: Pohang, KR

Pohang University of Science and Technology: Pohang, KR

Facebook: Boston, MA, US

Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory: Cambridge, MA, US

Education and qualifications (3)

Korea Advanced Institute of Science and Technology: Daejeon, Daejeon, KR

Korea Advanced Institute of Science and Technology: Daejeon, Daejeon, KR

Kwangwoon University: Nowon-gu, Seoul, KR

Works (50 of 61)

BEAF: Observing BEfore-AFter Changes to Evaluate Hallucination in Vision-Language Models

Learning-based Axial Video Motion Magnification

The Devil Is in the Details: Simple Remedies for Image-to-LiDAR Representation Learning

A unified framework for unsupervised action learning via global-to-local motion transformer

An Iterative Method for Unsupervised Robust Anomaly Detection Under Data Contamination

Factorized Multi-Resolution HashGrid for Efficient Neural Radiance Fields: Execution on Edge-Devices

Overcoming Client Data Deficiency in Federated Learning by Exploiting Unlabeled Data on the Server

Semi-Supervised Image Captioning by Adversarially Propagating Labeled Data

The devil in the details: simple and effective optical flow synthetic data generation

Uni-DVPS: Unified Model for Depth-Aware Video Panoptic Segmentation

Multi-stage adaptive rank statistic pruning for lightweight human 3D mesh recovery model

Computational discovery of microstructured composites with optimal stiffness-toughness trade-offs

ENInst: Enhancing weakly-supervised low-shot instance segmentation

Joint Video Super-Resolution and Frame Interpolation via Permutation Invariance

CLIP-Actor: Text-Driven Recommendation and Stylization for Animating Human Meshes

Cross-Attention of Disentangled Modalities for 3D Human Mesh Recovery with Transformers

HDR-Plenoxels: Self-Calibrating High Dynamic Range Radiance Fields

Lightweight Speaker Recognition in Poincaré Spaces

Neural-Network-Based Automated Synthesis of Transformer Matching Circuits for RF Amplifier Design

Dense Relational Image Captioning via Multi-Task Triple-Stream Networks

Learning to Localize Sound Sources in Visual Scenes: Analysis and Applications

Globally Optimal Inlier Set Maximization for Atlanta World Understanding

A Closed-Form Solution to Rotation Estimation for Structure from Small Motion

Gradient-based Camera Exposure Control for Outdoor Mobile Platforms

Learning-Based Video Motion Magnification

Robust and Globally Optimal Manhattan Frame Estimation in Near Real Time

Semantic soft segmentation

High-Fidelity Depth Upsampling Using the Self-Learning Framework

Contextually Customized Video Summaries Via Natural Language

Disjoint Multi-task Learning Between Heterogeneous Human-Centric Tasks

Fast Randomized Singular Value Thresholding for Low-Rank Optimization

Personalized Cinemagraphs Using Semantic Understanding and Collaborative Learning

Weakly- and Self-Supervised Learning for Content-Aware Deep Image Retargeting

A Pseudo-Bayesian algorithm for Robust PCA

Globally optimal manhattan frame estimation in real-time

Human body part classification from optical flow

Partial sum minimization of singular values in robust PCA: Algorithm and applications

Video-story composition via plot analysis

Globally Optimal Manhattan Frame Estimation in Real-Time

Video-Story Composition via Plot Analysis

A Multi-view Structured-Light System for Highly Accurate 3D Modeling

A two phase approach for pedestrian detection

An Autonomous Driving System for Unknown Environments Using a Unified Map

Fast randomized Singular Value Thresholding for Nuclear Norm Minimization

Line meets as-projective-as-possible image stitching with moving DLT

Robust high dynamic range imaging by rank minimization

Line assisted vision applications in structured environments

Balanced optical flow refinement by bidirectional constraint

Cost-aware depth map estimation for Lytro camera

Hierarchical 3D line restoration based on angular proximity in structured environments