Ph.D. Computer Science
2015 -- 2021
UC Berkeley, Berkeley, CA
Thesis Topic: Cinematic Editing for Casual Videos
B.S. Electrical and Computer Engineering
2011 -- 2015
Rice University, Houston, TX
Summa Cum Laude and Distinction in Research
The Berkeley Artificial Intelligence Research (BAIR)
Fall 2016 -- present
Graduate Student Mentor: Mentor undergrads from underrepresented groups on career and academic research.
Computer Vision for Global Challenges (@CVPR 2019)
Ambassador: Mentored CV4GC awardees to give presentation related to the developing world at CV4GC Workshop.
Research Intern, G-cam
Google, Mountain View, CA
Research Intern, Intelligent Systems Lab
Intel, Santa Clara, CA
Research Intern, Computational Photography
Facebook Research, Seattle, WA
Research Intern, Imagination Lab
Adobe Research, San Jose, CA
My research aims to build an editing system for casual videos to enable cinematic focus, perceptual image quality and lighting. My goal is to make casual videos, such as the kind taken by personal devices (e.g. smartphones), an effective and ubiquitous visual storytelling form.
I feel truly fortunate to have collaborated with these inspiring people from the industrial research labs (in chronicle order): Vladlen Koltun (Intel), Qifeng Chen (Intel), Kevin Matzen (Facebook), Joon-Young Lee (Adobe), Kalyan Sunkavalli (Adobe) , Zhaowen Wang (Adobe) .
Synthetic Defocus and Look-Ahead Autofocus for Casual Videography
Xuaner (Cecilia) Zhang, Kevin Matzen, Vivien Nguyen, Dillon Yao, You Zhang, Ren Ng SIGGRAPH 2019
Zoom to Learn, Learn to Zoom
Xuaner (Cecilia) Zhang, Qifeng Chen, Ren Ng, Vladlen Koltun CVPR 2019
Single Image Reflection Removal with Perceptual Losses
Xuaner (Cecilia) Zhang, Ren Ng, Qifeng Chen CVPR 2018
Photometric stabilization for fast-forward videos
Xuaner (Cecilia) Zhang, Zhaowen Wang, Kalyan Sunkavalli, Joon-Young Lee
OpenCV2 @ CUDA 9.0
CVPR2019 @ Long Beach
Mini Lecture @ Stereoscopic Perception
CVPR2018 @ Salt Lake City
Here are more projects I did for classes and for fun.Computer Graphics
In this project, we present a system for cross-domain similarity search that helps us with sketch-based 3D shape retrieval. Instead of using hand crafted features for searching, we propose our DeepSketch neural network that is built on Siamese network to learn features that are basis for later similarity search using K-nearest Neighbor (KNN). Here is a report on this project.
In this project, we built an Android APP for an AR food menu. We place a user-selected food image onto a plate in the real world space with a vision-based estimation of the correct homography. Here is a demo link.