🎤 Talks & Presentations

Delivered a talk on multilingual font generation using diffusion models, highlighting innovations in style fusion, structure preservation, and text-guided synthesis across Korean, Chinese, and Latin scripts.

Presented a poster on DK-Font, a text-conditioned diffusion model for generating high-fidelity Korean fonts from a single glyph. The presentation highlighted phonetic-aware content encoding, stroke preservation, and improvements over baseline methods like Diff-Font.

Delivered weekly lab talks on research progress, covering topics such as multilingual diffusion models, style-content disentanglement, and structured loss design for font generation.