Kun He
何坤 • Researcher at Meta Reality Labs • FirstnameLastname @ meta.com
Meta Reality Labs
Redmond, WA, USA
I am a research scientist at Meta Reality Labs, working on the future of interactions for Virtual Reality and Augmented Reality, using computer vision and machine learning techniques. Here’s my CV.
My current research interests include: human body/hand pose estimation and tracking, domain adaptation, and active learning.
I graduated from Boston University in 2018 with PhD and MSc degrees in computer science. My advisor was Professor Stan Sclaroff. My PhD dissertation focused on learning vector embeddings for visual data, which finds application in image retrieval and other related computer vision tasks. During grad school, I did research internships at Disney Research and Honda Research Institute, and I was a visiting student at Brown University. I hold a Bachelor’s degree in computer science from Zhejiang University, China.
news
Jun 2023 | AssemblyHands dataset will be featured in the 7th HANDS workshop at ICCV 2023: Task 1 - Egocentric 3D Hand Pose Estimation. |
---|---|
Feb 2023 | Paper accepted to CVPR 2023 (with Takehiko Ohkawa et al.): AssemblyHands: Towards Egocentric Activity Understanding via 3D Hand Pose Estimation. |
Oct 2022 | Paper accepted to WACV 2023 (with Qi Feng et al.): Rethinking the Data Annotation Process for Multi-view 3D Pose Estimation with Active Learning and Self-Training. |
Mar 2022 | Paper accepted at CVPR 2022 (with Fadime Sener et al.): Assembly101: A Large-Scale Multi-View Video Dataset for Understanding Procedural Activities |
Sep 2019 | Check out hand tracking on Oculus Quest! |
publications
conference papers
-
AssemblyHands: Towards Egocentric Activity Understanding via 3D Hand Pose EstimationIn Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023
-
Generalized Majorization-MinimizationIn Proc. International Conference on Machine Learning (ICML) 2019
-
Scale Resilient, Rotation Invariant Articulated Object MatchingIn Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2012
journal papers
-
Predicting Foreground Object Ambiguity And Efficiently Crowdsourcing the Segmentation(s)International Journal of Computer Vision (IJCV) 2018
theses
-
-
Stochastic Functional Descent for Learning Support Vector MachinesM.Sc. Thesis, Boston University, 2013