Kyungjune Baekمشاهده پروفایل
استادیار
Kyungjune Baek is an Assistant Professor in the Department of Software at Sejong University, South Korea, since March 2025. Previously, he served as an AI Research Engineer at Upstage (2023-2024) and ML Research Scientist at TwelveLabs (2024-2025). His educational background includes: Ph.D., Yonsei University (2023) B.S., Yonsei University (2018) Dr. Baek's research centers on generative models , synthetic datasets , self-supervised learning , and video/3D understanding . His work pioneers techniques for enhancing model robustness through privacy-preserving synthetic data generation, novel regularization methods, and efficient learning from unlabeled or noisy data sources, with significant applications in computer vision. His 2021-2022 publications reveal a cohesive research trajectory focused on improving deep learning reliability through data-centric innovations. Key contributions include synthetic data pretraining for GANs, self-distillation frameworks for noisy labels, and context-aware regularization techniques, all advancing state-of-the-art performance in visual recognition tasks while addressing critical challenges like data privacy and annotation scarcity. No scientific awards were documented in the source material. Available information indicates no formal advising relationships or grant funding details. His professional activities appear concentrated in academic research publication and industry R&D collaboration without mention of dedicated laboratory infrastructure or team leadership.