Cecilia Zhang

张璇儿

Ph.D. Candidate
EECS & Berkeley AI Research
University of California, Berkeley

Email: cecilia77 at berkeley dot edu
CV:     Github:     Flickr:     Instagram:

[FUN 6/2020] Photos of a recent trip to Yosemite, tried some new (unnatural) color grading.

Me
I am a CS PhD candidate at UC Berkeley, advised by Prof. Ren Ng. My research interests are in computational photography and machine learning.
------ Houston ---- [1,919.5 mi] ---- Berkeley ------
Prior to Berkeley, I got my B.S. from Rice University with a focus in computer vision under supervision of Prof. Ashok Veeraraghavan.

------- Beijing ---- [7,205 mi] ---- Houston -------
I was born and grew up in Beijing, an ancient city on its unique modernization path. I've shot a pile of photographs of my hometown, take a look at my Flickr albums!



Selected Publications ( Full Publications)

Portrait Shadow Manipulation
SIGGRAPH 2020 SIG 2020 paper figure
Synthetic Defocus and Look-Ahead Autofocus for Casual Videography
SIGGRAPH 2019 SIG 2019 paper figure

Education

berkeley eecs logo Ph.D. Computer Science 2015 -- 2020
UC Berkeley, Berkeley, CA

Thesis Topic: Bringing cinematic rendering into casual photos and videos

B.S. Electrical and Computer Engineering 2011 -- 2015 Rice ECE Logo
Rice University, Houston, TX

Summa Cum Laude and Distinction in Research


Mentorship / Teaching Experience

The Berkeley Artificial Intelligence Research (BAIR) Fa16 -- Sp20
Graduate Student Mentor: Mentor undergrads from underrepresented groups on career and academic research.

TechX Academy Bootcamp Summer 2019
Instructor: Designed and taught a 10-day course on "Intro to Computer Vision".

Computer Vision for Global Challenges (@CVPR 2019) Summer 2019
Ambassador: Mentored CV4GC awardees to give presentation related to the developing world at CV4GC Workshop.

CS184/284A Computer Graphics & Imaging Fa16, Sp18
Graduate Student Instructor: Taught weekly discussion sessions.


Industrial Experience

Research Intern Summer 2019 Google Logo
Google Research (Gcam), Mountain View, CA

Research Intern, Intelligent Systems Lab Summer 2018 Intel Logo
Intel, Santa Clara, CA

Research Intern, Computational Photography Summer 2017 Facebook Logo
Facebook Research, Seattle, WA

Research Intern, Imagination Lab Summer 2016 Adobe Logo
Adobe Research, San Jose, CA

Research

My research aims to build an editing system for casually-taken photos and videos to enable cinematic focus, perceptual image quality and lighting. My goal is to make casual photos and 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): David E. Jacobs (Google Research), Jonathan T. Barron (Google Research), Yun-Ta Tsai (Google Research), Vladlen Koltun (Intel), Qifeng Chen (Intel), Kevin Matzen (Facebook), Joon-Young Lee (Adobe), Kalyan Sunkavalli (Adobe) , Zhaowen Wang (Adobe) .


Conference Papers

Portrait Shadow Manipulation
Xuaner (Cecilia) Zhang, Jonathan T. Barron, Yun-Ta Tsai, Rohit Pandey, Xiuming Zhang, Ren Ng, David E. Jacobs
SIGGRAPH 2020  

SIG 2020 paper figure


Synthetic Defocus and Look-Ahead Autofocus for Casual Videography
Xuaner (Cecilia) Zhang, Kevin Matzen, Vivien Nguyen, Dillon Yao, You Zhang, Ren Ng
SIGGRAPH 2019  

SIG 2019 paper figure


Zoom to Learn, Learn to Zoom
Xuaner (Cecilia) Zhang, Qifeng Chen, Ren Ng, Vladlen Koltun
CVPR 2019  

cvpr 2019 paper figure


Single Image Reflection Removal with Perceptual Losses
Xuaner (Cecilia) Zhang, Ren Ng, Qifeng Chen
CVPR 2018  

cvpr 2018 paper figure


Photometric stabilization for fast-forward videos
Xuaner (Cecilia) Zhang, Zhaowen Wang, Kalyan Sunkavalli, Joon-Young Lee
PG 2017  

PG 2017 paper figure


Virtual Fitting: real-time garment simulation for online shopping
Xuaner (Cecilia) Zhang, Lam Yuk Wong
CASA 2014  


Conference Posters

Free Your Eyes: Retinal Image Deblurring Display with Enlarged Viewing Zone
Xuaner (Cecilia) Zhang, Jinkyu Kim, Brian Barsky, Laura Waller, Ren Ng
ICCP 2015, Poster  

iccp16 Poster

Blogs

Here I write random things about the academic and life. Hope you can find some of the pieces resonating and helpful. Send me an email for any suggestion or comment!

05-22-2020: Job Interview Resources (targeting PhD grads)

Job Interview Resources @ CV/ML

Curated lists of review resources for Computer Vision and Machine Learning, targeting PhD grads and research interviews.
READ MORE


11-04-2019: Asian Dean Rising Star

Asian Dean Rising Star Workshop @ Seoul

A 2-day workshop that taught me how challenging it is to become a women faculty, but also gave me full confidence to get there.
READ MORE


09-16-2019: OpenCV2 with CUDA 9.0

OpenCV2 @ CUDA 9.0

Hours of battle with OpenCV2 compilation with CUDA 9.0
READ MORE


06-22-2019: CVPR 2019

CVPR2019 @ Long Beach

8K+ attendees and 1294 papers, I tried hard to sort myself out to enjoy it. It was exhausting, but worthwhile and fun.
READ MORE


03-18-2019: The cardboard effect explained

Mini Lecture @ Stereoscopic Perception

"Only when no artifacts are visible would stereoscopic imagery be preferred." Our visual system is very demanding about 3D contents.
READ MORE


06-23-2018: CVPR 2018

CVPR2018 @ Salt Lake City

5 days, 979+ papers and 6000+ attendees. While being a bit overwhelmed, I wrote down things I have experienced, learned and reflected.
READ MORE

More Projects (& Pretty Images)

Here are projects I did for fun (or for fun classes).

Computer Graphics
Below are some projects I did throughout this graphics course, covering a variety of cool topics such as rasterizing, texture mapping, geometry processing, ray tracing, etc. Please click on each project to view high resolution renderings.

Computational Photography
  
DeepSketch: Sketch-based 3D shape Retrieval

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 upon the Siamese network to learn features for KNN search. Here is a report on this project.


  
Cinematic Color and Colorization

It is well known that cinematographers may carefully select colorsto provoke an emotional response, provide visual clarity, or serve anarrative purpose. We also have the sense, however, that the simple set of colors used in cinema is distinguishable both from the colors of casual videos and among different cinematic genres. In this project, we analyze how color is used in cinema and exploremethods to transfer color between cinematic and casual videos. Here is the video and write-up for this project.


The rest ancient, naïve but free ideas
  
AR Menu

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.


Virtual Drum Kit

Our virtual drum design simulates the experience of playing drums. The system includes: video processing to track the position of the drum stick in real time; audio processing to detect the ‘hit’ moment of the drum stick to generate corresponding drum sounds.


More About Cecilia

Basically, I like everything visual.


Digital Photography --- I love taking photos (click me!), like street-views, nature with a structure, objects with granularity and people when they are relaxed. Nowadays I'm using Fujifilm XT-20 and occasionally just my pocket iPhone X. I believe in the content and moment more than the device. I use photographs to retrieve memories and to capture moments that's passing by too fast to be seen by our eyes.



Film Photography --- When I do film photography, I put myself in a gray-scale mode to focus on lighting and shadow rather than colors. I enjoy spending hours in the darkroom and developing the film with classic dodge and burn. I found it interesting that images often turned out to be different from what you think you originally have captured. I like this feeling of a delayed gratification, and a sense of surprise when you see something different from what your brain recorded or rendered.



Travel --- As I visit more places, I meet more people, experience more interesting moments and get to take more photos. My favorite places I've visited so far are Bremen, Interlaken and the summer version of Seattle.



Coffee --- I'm a coffee-addict. More precisely, I'm a cafe-addict. It's interesting to discover the "personality" of each cafe, its baristas, its coffee machine preference and the different taste of their 'house coffee'.



Street-wandering --- Street-view tells a lot about a city. If you look at the pictures below without hovering to see their labels, you might already had a reasonable guess about its geo-location. Street-views also connect with the local, and help me understand a city better.


My Significant "Since ..." Moments

Since Rice


I miss Rice often, it's the place where I got to discover my current research interests and met many fellow friends who are now great researchers and talented engineers. I also met my favorite professors who brought me into the field of image processing and computational photography, who guided me to see the interdisciplinary beauty of computation and cameras. Among the campuses I've visited (>20 schools), Rice is the pretties, with no doubt.
Since RDFZ


I spent my 6 years of middle + high school in RDFZ (aka. 人大附中), and had the best three years of my life with class 12. I met some of my most cherished people, lifelong friends and also my significant one. It would forever be a part of my unforgettable memories. We are spread around the world, but would gather together for reunion almost every year even after 7 years of graduation.
Since Yoo


Yoo and me met in high school, we were in the same class (we actually first met in elementary school but didn't know each other at the time). He just graduated with a MFA in film-making and is now a freelancer director. He owns his studio back in China and is pursuing his dream in telling great stories in the format of films to others. Take a look at his SHOWREEL 2017!
Since 米粒儿


米粒儿 joined the family in pure black when she was a few months old. Now she's more mixed gray-ish. She loves chicken flavor snacks, and hates wearing clothes or shoes that we would think look cute on her. She is very smart, all the smart toys or puzzles we bought her would be completely solved and abandoned by her in a day or two. My mom is her favorite, than goes my Dad, and then Yoo and very lastly it's me. I hope I could see her more frequently and let her curl up on my lap.