Inhyuk Song
Department of Artificial Intelligence, Kyung Hee University, Republic of Korea
cherished990204@khu.ac.kr · GitHub · LinkedIn
Research interests
Computational models of human perception and action; tactile perception; adaptive physical feedback; personalized interaction design.
Education
Kyung Hee University — Master’s studies in Artificial Intelligence
September 2025 – August 2027 (expected) · GPA: 4.2 / 4.3
Adviser: Seungjae Oh
Kyung Hee University — Bachelor’s degree in Software Convergence
March 2018 – August 2025
College of Software Convergence · Game Contents Track
Military service: Republic of Korea Air Force, July 2020 – April 2022.
Research experience
ITEM Lab, Kyung Hee University
Graduate Student Researcher: September 2025 – present
Undergraduate Researcher: September 2024 – August 2025
- Tactile perception modeling: Developed computational models of tactile perception and evaluated their predictions against psychophysical data.
- Adaptive mid-air feedback: Led the design, implementation, and exploratory evaluation of a vibrotactile feedback system that adapts to pressing behavior.
- SwapSense: Conducted multidimensional scaling (MDS) analysis of perceptual relationships among interchangeable haptic modules.
- Current direction: Developing a research approach connecting perceptual modeling, biomechanical simulation, and reinforcement learning for personalized physical interaction.
Accepted conference paper
Minha Jeon, Yeseo Park, Inhyuk Song, Yilong Lin, Seungwoo Je, and Seungjae Oh. SwapSense: Reusable Vision-Based Contact Force Sensing Across Swappable Passive Haptic Modules. ACM Symposium on User Interface Software and Technology (UIST 2026). Accepted.
Contribution: multidimensional scaling analysis.
Conference presentations and posters
Inhyuk Song, Hee-Seung Moon, and Seungjae Oh. Comparing Peripheral Signal Representations for Predicting Perceived Intensity of Vibrotactile Stimuli. Korean Haptics Conference 2026. Presented.
Inhyuk Song and Seungjae Oh. Spacewalk Locomotion Technique and Tactile Feedback for Enhancing VR Experience while Lying Down. Korean Haptics Conference 2024. Accepted for poster presentation.
Selected projects
Inhyuk Song and Kihyun Park. Making Mid-air Button Contact Perceptible: Adaptive Feedback Model for Enhancing Interaction Saliency in VR. WHC 2025 Student Innovation Challenge.
Led project design, implementation, experimental design, and analysis. Developed adaptive vibrotactile feedback for mid-air button interaction and evaluated pressing behavior in an exploratory user study.
Undergraduate team projects: Hand-tracked VR locomotion (March–June 2024); RGB-D scene reconstruction and proximity-aware VR guidance (September–December 2023). Project details.
Teaching experience
Teaching Assistant — Game Engineering (SWCON314)
Kyung Hee University · 2026 – present
- Lead practical sessions as the course’s sole teaching assistant.
- Course topics include kinematics and numerical optimization in Unreal Engine, hand-motion prediction, and neural-network inference in Unity.
Teaching Assistant — Game Interactive Technology (SWCON312)
Kyung Hee University · Spring 2026
- Supported practical instruction in Unity input systems, custom HID integration, VR controller input, and hand tracking and pose detection.
Teaching Assistant — Introduction to Game Programming (SWCON211)
Kyung Hee University · Fall 2025
- Supported Python/Pygame labs on motion simulation, collision detection (AABB/SAT), impulse-based collision response, and PyOpenGL rendering.
Academic service
- Student Volunteer, IEEE World Haptics Conference (WHC 2025).
Courses and workshops
- Introduction to Computational Cognitive Modeling — CHI 2025 course participant.
- Mind & Motion: Opportunities and Applications of Integrating Biomechanics and Cognitive Models in HCI — UIST 2025 workshop participant.
Research methods
Computational perceptual modeling; tactile-afferent simulation; representation learning; model ablation; psychophysical data analysis; multidimensional scaling; VR prototyping; experimental design and analysis.