Journals
Gayeon Kim, Yugyeong Hong, Jong Yoon Lim, Trevor Gee, Bruce A. MacDonald, Ho Seok Ahn - “Emotional Talking Face Generation with a Single Image.”
Proceedings of the 21st IEEE International Conference on Ubiquitous Robots (UR 2024). [paper]
Also, Poster presentation at the Australasian Conference on Robotics and Automation (ACRA 2023).
C. N. Cho, Y. Hwang, S. G. Yang, H. S. Jeong, G. Y. Kim, J. H. Hwang, T. K. Kim. “VLM-Based Subtask Generation for Long-Horizon Tasks.” Under review for publication in the Korean Robotics Society Conference (KRoC 2026) proceedings.
Conferences
H. Y. Chung, H. S. Jung, G. Y. Kim, M. S. Oh, K. H. Kim, C. N. Cho, J. H. Hwang, T. K. Kim. “Marker-Based Pose Estimation for Robotic Bolting in Reconfigurable Manufacturing.” Accepted for presentation at the Korean Society of Manufacturing Technology Engineers (KSMTE 2025). Proceedings submission under review.
Emotional Talking Face Generation — University of Auckland
Developed a deep neural network pipeline that generates emotional talking-face videos from a single image and text input. Combined audio–visual feature extraction, GAN-based image synthesis, and head-pose dynamics for realistic emotional expression. Outperformed state-of-the-art models such as Wav2Lip and PC-AVS.
Robotic Assembly and Simulation Framework — KETI
Developing ROS2–Gazebo simulation environments and real-world control systems for multi-robot assembly using AprilTag-based visual localization and OPC UA communication. Enables precise calibration and real-time motion synchronization between simulated and physical systems.
Autonomous Mobile Robot Navigation — University of Auckland
Built a SLAM-enabled mobile robot integrating depth-camera data and odometry for real-time mapping and obstacle avoidance. Implemented probabilistic localization algorithms and autonomous path planning in ROS.
Korea Electronics Technology Institute (KETI) - Research Institute
Robotics Researcher
April 2025 – Present
Developing robotic assembly systems for high-precision automation, integrating perception, calibration, and control with Universal Robots UR10 and Denso robotic arms.
Performed hand–eye calibration and coordinate alignment between robot, tool, and camera frames to achieve sub-millimetre accuracy in assembly operations.
Introduced a Remote Center Compliance (RCC) device to improve insertion precision, providing a simpler and more robust alternative to additional vision compensation.
Built a ROS2–Gazebo digital twin connected via OPC UA for safe, hardware-independent testing, and refined simulation stability through physics and collision tuning.
Validated system precision using a laser tracker to assess positional repeatability at fixed reference points.
Conduct live robotic demonstrations for visitors and collaborators, ensuring safe and reliable operation during presentations.
Development of a humanoid equipped with a multimodal AI foundation-model–based high-sensitivity robotic hand
Fonterra Co-operative Group Limited — IT Intern, Intelligent Automation Team
December 2023 – March 2024
Applied machine learning and data analytics to business optimisation problems.
Built and deployed a Copilot Studio chatbot integrated with Power Automate and Co-op GPT, reducing post-project review time by 60 %.
Developed Python-based predictive models for ROI and product-trend forecasting and interactive Power BI dashboards for decision support.