Paper 'Inference-Time Diffusion Model Distillation' accepted at ICCV 2025
Paper 'VideoGuide' accepted at CVPR 2025
'TReg' selected as Spotlight at ICLR 2025; both 'TReg' and 'CFG++' accepted at ICLR 2025
'CFG++' awarded 2nd place in Signal Processing at the 31st Samsung Humantech Paper Award
'DreamMakeup' won 2025 CES Innovation Award (collaborated with AmorePacific)
Paper 'SMA' accepted at AAAI 2025
'VMC' presented as oral at KCCV 2024 and finalist for Qualcomm Innovation Fellowship Award 2024
Papers 'Dreamsampler' and 'DreamMotion' accepted at ECCV 2024
Three papers ('Self-supervised debiasing', 'VMC', 'CDS') accepted at CVPR 2024
'EBCA' received 3rd place in Signal Processing at the 30th Samsung Humantech Paper Award
'EBCA' presented as a poster at NeurIPS 2023
Background
Final-year Ph.D. student in Bio and Brain Engineering (BBE) at KAIST
Aspires to become a tech-based leader in AI
Research interests: (a) improving representation learning; (b) leveraging intermediate representations for controlling AI systems
Recent work focuses on controlling diffusion generative processes via latent representations or denoised estimates to enhance conditional diffusion sampling and multi-modal alignment
Interested in developing robust representations to handle distribution shifts