Seong-Eun Hong

Hello, I'm Seong-Eun Hong

Ph.D. Student @ Korea University

Exploring the intersection of Computer Graphics, Motion Synthesis, and Deep Learning

About Me

I am a Graphics Researcher and Ph.D. Student at Korea University. My research focuses on understanding and imitating human behaviors through Computer Graphics and Deep Learning.

Research Interests

Computer Graphics Natural Language Processing Virtual Agents Crowd Simulation Motion Synthesis Deep Learning
0
Publications
0
Awards
0
Projects

Education

2024.03 - 2030.02 (Expected)

Korea University

Doctor of Philosophy - Ph.D. Student, Computer Science

Research Interest: Computer Graphics

2018.03 - 2023.02

Kyung Hee University

Bachelor, Software Convergence

GPA: 4.04/4.5
Acquired Skills: Deep Learning, Natural Language Processing (NLP)

Experience

2023.09 - 2024.02 (6 mos)

IIIXR Lab

Intern · On-site

2023.03 - 2023.08 (6 mos)

TmaxA&C

NLP Researcher · Full-time

Pangyo, South Korea · On-site

Publications

2026
Event-T2M

Event-T2M: Event-level Conditioning for Complex Text-to-Motion Synthesis

Seong-Eun Hong*, Jaeyoung Seon*, Juyeong Hwang, JongHwan Shin, and HyeongYeop Kang

ICLR 2026 Conference Paper

Rio de Janeiro, Brazil | Apr. 23-27, 2026

A novel approach for event-level conditioning in text-to-motion synthesis, enabling complex and nuanced motion generation from text descriptions.

ORACLE

ORACLE: Orchestrate NPC Daily Activities using Contrastive Learning with Transformer-CVAE

Seong-Eun Hong, Juyeong Hwang, Ryunha Lee, and HyeongYeop Kang

CVM 2026 Conference Paper

Seoul, South Korea | Apr. 2026

A framework for orchestrating NPC daily activities using contrastive learning combined with Transformer-CVAE architecture for realistic virtual agent behaviors.

BiPO

BiPO: Bidirectional Partial Occlusion Network for Text-to-Motion Synthesis

Seong-Eun Hong, Soobin Lim, Juyeong Hwang, Minwook Chang, and HyeongYeop Kang

WACV 2026 Conference Paper Oral Presentation

Tucson, Arizona, USA | Mar. 6-10, 2026

A bidirectional partial occlusion network designed for improved text-to-motion synthesis, achieving state-of-the-art performance and selected for oral presentation.

2025
Virtual Agent Head Rotation

How Does a Virtual Agent Decide Where to Look? Symbolic Cognitive Reasoning for Embodied Head Rotation

Juyeong Hwang*, Seong-Eun Hong*, Jaeyoung Seon, and HyeongYeop Kang

SIGGRAPH Asia 2025 Conference Paper Oral Presentation

Hong Kong | Dec. 15-18, 2025

A novel framework for symbolic cognitive reasoning enabling virtual agents to make natural head rotation decisions in embodied environments.

2022
AACL Paper

Post-Training with Interrogative Sentences for Enhancing BART-based Korean Question Generator

Gyu-Min Park, Seong-Eun Hong, and Seong-Bae Park

AACL-IJCNLP 2022 Conference Paper

Taipei City, Taiwan | Nov. 20-23, 2022

Proposed a post-training method using interrogative sentences to improve the performance of BART-based Korean question generation models.

Patents

2025.10.16

Method and System for Determining Head Rotation Based on Agent's Gaze

App. No: 10-2025-0149262

2024.04.24

System and Method for Generating Daily Activity Plans for Virtual Agents

App. No: 10-2025-0155836

2022.11.18

Additional Training Technique for Language Models for Query Generation

App. No: 10-2022-0155485

Projects

Text-to-Motion Generation

Joint research with NCSOFT. Developing high-fidelity motion synthesis models capable of generating complex interactions from natural language descriptions.

Diffusion Models PyTorch NLP

Uncertainty-Aware Self-Growing Agent

Joint research with ETRI. Developing agent technology that self-improves by recognizing uncertainty and asking questions to resolve ambiguity.

Question Generation Fine-tuning Post-training

Awards & Honors

2025

Best Paper Award

KCGS 2025

2022

Best Paper Award

KSC 2022

2022

Excellence Presentation Award

KCC 2022

2022.11

Grand Prize

SW Festival · Kyung Hee University

2022.11

Excellence Award

Khuthon 2021 · Kyung Hee University

2022.11

Popularity Award

Khuthon 2021 · Kyung Hee University

Get in Touch

I'm always open to discussing new research collaborations, interesting projects, or opportunities in graphics and motion synthesis.