kimmanjin Goto Github PK

I'm Manjin Kim, a Ph.D student in Pohang University of Science and Technology (POSTECH), South Korea. I am a member of the computer vision lab at POSTECH, under supervision of Professor Minsu Cho. My research interest is in learning video representation and its applications.

• Video representation learning
• Motion feature learning
• Multi-modal learning

• Learning correlation structures for vision transformers [project page] [code]
  Manjin Kim, Paul Hongsuck Seo*, Cordelia Schmid, and Minsu Cho* (* corresponding authors), CVPR 2024.
• Future transformer for long-term action anticipation [project page] [code]
  Dayoung Gong, Joonseok Lee, Manjin Kim, Seongjong Ha, and Minsu Cho, CVPR 2022.
• Relational self-attention: what's missing in attention for video understanding [project page] [code]
  Manjin Kim*, Heeseung Kwon*, Chunyu Wang, Suha Kwak, and Minsu Cho (* equal contribution), NeurIPS 2021.
• Learning self-similarity in space and time as as generalized motion for video action recognition [project page] [code]
  Heeseung Kwon*, Manjin Kim*, Suha Kwak, and Minsu Cho (* equal contribution), ICCV 2021.
• MotionSqueeze: neural motion feature learning for video understanding [project page] [code]
  Heeseung Kwon, Manjin Kim, Suha Kwak, and Minsu Cho, ECCV 2020.

• Student Researcher, Google Research, France (Jul. 2022 - Jan. 2023)
  • Developed a multimodal long-form video captioning system.
  • Host: Paul Hongsuck Seo
• Research Intern, Microsoft Research Asia (MSRA), remote (Dec. 2020 - June. 2021)
  • Developed a dynamic neural feature transform method, called Relational Self-Attention.
  • Mentor: Chunyu Wang
• Research Intern, LG CNS, Korea (Jun. 2018 - Aug. 2018)
  • Developed a video data augmentation system using CycleGAN.

• CV: CV
• e-mail: [email protected]
• google scholar: https://scholar.google.com/citations?user=kqddtlwAAAAJ&hl=en&oi=ao