University of California, Berkeley
Cecilia Zhang is a 4th year PhD candidate in Computer Science at UC Berkeley advised by Ren Ng. She works on computational photography, computer vision and machine learning. Her thesis is on cinematic capturing and rendering for casual videos to enable effective visual content creation from personal devices. She received her Bachelors degree in Electrical and Computer Engineering from Rice University in 2015, awarded with Summa Cum Laude and Distinction in Research. She is a member of Women in Computer Science at UC Berkeley to help women graduate students build up leadership and professional skills.
Cinema videos are characterized by shallow depth of field that controls the viewer’s gaze, high-quality video frames and carefully designed lighting. Such cinematic characteristics can only be achieved because of scripted stories and a large, professional crew with large film cameras. Casual videos do not have a screenplay and are captured using regular cameras by regular people. Quality of casual videos is constrained by the size of imaging sensors and the challenge of delivering meaningful focus and lighting. My work addresses video focus, lighting and image quality using image-based and physically-based rendering, machine learning and scene understanding in computer vision, to make some of the fundamental challenges in casual video enhancement tractable for the first time. My vision is to narrow the gap between casual and cinematic content creation, opening up opportunities for more storytellers to share perspectives and instill values via high-quality casual contents.