[세미나] Seungjun Nah Ph.D. (Research Scientist at NVIDIA)
페이지 정보
작성자 최고관리자 댓글 조회 작성일 23-10-10 15:14본문
Period : October 18, 2023 Wednesday 14:00~
Location : Engineering Hall 4, Room D504
Building Foundational AI Models for Text-Based Visual Content Generation
Abstract
Generative AI has become extremely popular in commercial applications of computer vision. Despite the long history of research on generative models, the emergence of the generative AI era has only been backed by recent large-scale diffusion models. These applications are mainly focused on language-based content creation. This talk will cover my work on large-scale text-to-image and text-to-video generation models that are developed in series.
Generative AI has become extremely popular in commercial applications of computer vision. Despite the long history of research on generative models, the emergence of the generative AI era has only been backed by recent large-scale diffusion models. These applications are mainly focused on language-based content creation. This talk will cover my work on large-scale text-to-image and text-to-video generation models that are developed in series.
Bio
Seungjun Nah is a research scientist at NVIDIA. He received his Ph.D. with distinguished dissertation award and B.S. from the Department of ECE at SNU, Korea. Prior to joining NVIDIA, he worked as a postdoctoral researcher at SNU, a guest scientist at the Max-Planck-Institute for Intelligent Systems, and a research intern at Microsoft. He pursues high-fidelity visual content generation and the computational efficiency of such methods. His recent research interests include text-based image and video generation, large-scale diffusion models, and image restoration problems. He is the winner of the NTIRE 2017 image super-resolution challenge. He co-organized several international workshops and hosted image/video restoration competitions at NTIRE 2019-2021, AIM 2019-2020, and AI4CC 2022-2023. He has served as an outstanding reviewer at CVPR 2021, ICCV 2019, 2021, and ECCV 2020, 2022 conferences.
Seungjun Nah is a research scientist at NVIDIA. He received his Ph.D. with distinguished dissertation award and B.S. from the Department of ECE at SNU, Korea. Prior to joining NVIDIA, he worked as a postdoctoral researcher at SNU, a guest scientist at the Max-Planck-Institute for Intelligent Systems, and a research intern at Microsoft. He pursues high-fidelity visual content generation and the computational efficiency of such methods. His recent research interests include text-based image and video generation, large-scale diffusion models, and image restoration problems. He is the winner of the NTIRE 2017 image super-resolution challenge. He co-organized several international workshops and hosted image/video restoration competitions at NTIRE 2019-2021, AIM 2019-2020, and AI4CC 2022-2023. He has served as an outstanding reviewer at CVPR 2021, ICCV 2019, 2021, and ECCV 2020, 2022 conferences.
댓글목록
등록된 댓글이 없습니다.
