Kris M. Kitani
Also known as K.M. Kitani, Kris Kitani
Kris Kitani is a computer-vision researcher at Carnegie Mellon University whose work models human activity from first-person video and supports hands-on learning. He earned a BS in electrical engineering at the University of Southern California, then MS and PhD degrees at the University of Tokyo, and held research posts in Tokyo and at CMU, where he directed the MSCV program. His papers modeled activities from first-person video and applied inverse reinforcement learning to forecast goals. He coauthored Ego4D and helped develop Origami Sensei; his work received a Marr Prize honorable mention at ICCV 2017 and best paper honorable mentions at CHI 2017 and 2020. He is an associate research professor in CMU's Robotics Institute and co-director of its Extended Reality Technology Center, and he consults for Roblox in 2026.
- Badges
Timeline
-
2000–2003
A USC degree and KLA-Tencor
He earns a BS in electrical engineering at the University of Southern California in 2000 and works as an engineer at KLA-Tencor from 2000 to 2003.
-
2005–2011
Tokyo degrees and professorship
He earns an MS in 2005 and a PhD in 2008 at the University of Tokyo, then works as an assistant professor at UEC Tokyo from 2008 to 2011 with Hideki Koike.
-
2011–2022
CMU research and program leadership
He joins CMU as a postdoctoral research fellow in 2011 with Martial Hebert and Drew Bagnell, becomes a systems scientist in 2013, and an assistant research professor in 2016. He directs CMU's MSCV program from 2018 to 2022.
-
2016
First-person activity forecasting
At ACCV 2016, he coauthors Long-Term Activity Forecasting using First-Person Vision, which applies deep reinforcement learning to predict full activities from egocentric video. He also coauthors First-Person Activity Forecasting with Online Inverse Reinforcement Learning; its DARKO model forecasts goals and semantic states from first-person observations.
-
2017–2020
Conference paper honors
His work earns a Marr Prize honorable mention at ICCV 2017, plus best paper honorable mentions at CHI 2017 and CHI 2020.
-
2022
The Ego4D dataset
In 2022, he coauthors Ego4D, a CVPR dataset and benchmark for egocentric video that contains 3,670 hours of daily-life footage from 931 camera wearers across 74 worldwide locations.
-
2023–2025
Meta research and Origami Sensei
From 2023 to 2025, he works as an AI Research Scientist at Meta in RL, FAIR and MSL. In 2025, he leads the development of Origami Sensei with Dina El-Zanfaly, a mixed-reality AI system that uses computer vision and a projector to help beginners create origami; the team presents the research at CHI 2025.
-
As of October 2026
Professor and XRTC co-director
He is an associate research professor in CMU's Robotics Institute and co-director of the Extended Reality Technology Center. In 2026, he consults for Roblox.
Career
- Since 2011Carnegie Mellon UniversityPostdoctoral Research Fellow, Systems Scientist
- 2023–2025MetaAI Research Scientist
- 2023–2025FAIR (Meta)AI Research Scientist
- 2023–2025Meta Superintelligence LabsAI Research Scientist
- 2008–2011University of Electro-CommunicationsAssistant Professor
- 2000–2003KLA-TencorEngineer
Education
- 2008University of TokyoPhD
- 2005University of TokyoMS
- 2000University of Southern CaliforniaBS
Sources and credits
Photo: We have no published photo of Kris M. Kitani yet. Send us a free-licensed photo.
- Kris Kitani PhD
- Long-Term Activity Forecasting using First-Person Vision
- First-Person Activity Forecasting with Online Inverse Reinforcement Learning
- Ego4D: Around the World in 3,000 Hours of Egocentric Video
- Mixed-Reality AI Assistant Teaches Hands-On Skills
- CMU Robotics Expertise and Facilities Attract IROS 2026 Conference Attendees
- Kris M. Kitani - The Robotics Institute - CMU



