Seunghyeon Seo

Seunghyeon Seo (서승현) 서울대학교 협동과정 인공지능전공 석박사통합과정 (2021.03 ~ ) 서울대학교 농경제사회학부 지역정보전공/인문데이터과학(연계전공) 학사 졸업 (2021.02) Tel: (+82)31-888-9579 e-mail: zzzlssh@snu.ac.kr Research Page Google Scholar LinkedIn Github

Research Interests

Deep Learning, Computer Vision, Human Pose Estimation, Neural Rendering

Work Experience

• Jun. 2020 ~ Nov. 2020:

  Research Assistant at ThinkforBL Consulting Group, Korea.

• Sep. 2019 ~ Feb. 2020:

  Intern at the Committee on World Food Security, Food and Agriculture Organization of the United Nations (FAO), Italy.

Education

• Mar. 2021 ~ Present:

  M.S./Ph.D. Student, Interdisciplinary Program in Artificial Intelligence, College of Engineering, Seoul National University, Korea.

• Feb. 2021:

  B.S., Regional Information Major / Data Science for the Humanities (Combined Minor), Seoul National University, Korea.

• Jan. 2019 ~ Jun. 2019:

  Exchange Student, Economics / International Relations, Sciences Po (Paris Campus), France.

Publications

International Conferences

• Donghoon Han*, Seunghyeon Seo*, Eunhwan Park, SeongUk Nam, Nojun Kwak, "Unleash the Potential of CLIP for Video Highlight Detection", * indicates equal contribution, CVPR 2024 Workshop on Efficient Large Vision Models.
• Yeonjin Chang, Yearim Kim, Seunghyeon Seo, Jung Yi, Nojun Kwak, "Fast Sun-aligned Outdoor Scene Relighting based on TensoRF", WACV 2024.
• Donghoon Han, Seunghyeon Seo, DongHyeon Jeon, Jiho Jang, Chaerin Kong, Nojun Kwak, "ConcatPlexer: Additional Dim1 Batching for Faster ViTs", NeurIPS 2023 Workshop on Advancing Neural Network Training. (Oral)
• Seunghyeon Seo, Yeonjin Chang, Nojun Kwak, "FlipNeRF: Flipped Reflection Rays for Few-shot Novel View Synthesis", ICCV 2023. (Project Page)
• Seunghyeon Seo, Jaeyoung Yoo, Jihye Hwang, Nojun Kwak, "MDPose: Real-Time Multi-Person Pose Estimation via Mixture Density Model", UAI 2023.
• Jaeyoung Yoo*, Hojun Lee*, Seunghyeon Seo, Inseop Chung, Nojun Kwak, "End-to-End Multi-Object Detection with a Regularized Mixture Model", * indicates equal contribution, ICML 2023. (Code)
• Seunghyeon Seo, Donghoon Han*, Yeonjin Chang*, Nojun Kwak, "MixNeRF: Modeling a Ray with Mixture Density for Novel View Synthesis from Sparse Inputs", * indicates equal contribution, CVPR 2023. (Project Page)
• Jongmok Kim, Jooyoung Jang, Seunghyeon Seo, Jisoo Jeong, Jongkeun Na, Nojun Kwak, "MUM: Mix Image Tiles and UnMix Feature Tiles for Semi-Supervised Object Detection", CVPR 2022. (Code)
• Kyuewang Lee*, Inseop Chung*, Daeho Um, Jaeseok Choi, Yeji Song, Seunghyeon Seo, Nojun Kwak, Jin Young Choi, "Multi-modal Object Detection, Tracking, and Action Classification for Unmanned Outdoor Surveillance Robots", * indicates equal contribution, ICCAS 2021.

Under Review

• Seunghyeon Seo, Yeonjin Chang, Jayeon Yoo, Seungwoo Lee, Hojun Lee, Nojun Kwak, "HourglassNeRF: Casting an Hourglass as a Bundle of Rays for Few-shot Neural Rendering", arXiv, Mar. 2024.

Projects

• Artificial intelligence research about cross-modal dialogue modeling for one-on-one multi-modal interactions, May. 2022 ~ Jul. 2023
• Development of real-time multi-camera object tracking and identification technology, Jun. 2021 ~ Dec. 2021
• Development of multimodal sensor-based intelligent systems for outdoor surveillance robots, Jan. 2021 ~ Aug. 2021