Massachusetts Institute of Technology

The University of Hong Kong

Sr Machine Learning Researcher at Vimeo

New York, New York, United States

Vimeo

Senior Machine Learning Researcher

Designed and prototyped a mobile-AR shared-experience storytelling system, where one presenter can tell a story in realtime, using gestures and body positions to manipulate virtual objects. 3 and more viewers can watch simultaneously with mobile devices. ARKit, Unity, C#

Adobe

Research Engineer

- Built a novel system in Unity that automatically generates animated movies based on personal data for mood and depression intervention. Ran human subject tests involving 50 users. Published in Proceedings of the 2018 ACM CHI Conference.
-  Developed a personalized deep learning algorithm with TensorFlow to predict subjective pain from facial expressions that is 15% more accurate than the state of the art, and the first to visualize individual expressions of pain. Published in Proceedings of Machine Learning Research.
-  Worked in a team of 3 to build a mobile game in Unity that increases daily self-report compliance by 1.85 times, compared to e-forms. Published in Journal of Medical Internet Research Preprints, 2018.

MIT Media Lab

Research Assistant

Worked on a deep learning image segmentation algorithm with C++ and Caffe. Improved pixel-level segmentation accuracy by adding a higher order term to the conditional random fields energy.

SenseTime Group Limited

Research Scientist Intern

- To synchronize videos of similar motions, helped develop a belief propagation type of optimization with Matlab, C++ and OpenCV, which significantly outperforms dynamic time warping (DTW) in synthetic data tests.

Fengjiao Peng