Developed a novel, cost-effective lighting representation called Envmap++ for accurate reconstruction of glossy objects in indoor environments.
Conducted research on improving the fidelity of glossy object reconstruction under complex indoor illumination conditions.
Link to our arXiv submission:
Research Intern
Meta Reality Labs
Explored baking artifacts in material reconstruction using inverse rendering, proposing a potential method to address these challenges.
Collaborated on developing a hybrid pipeline that combines NeRF with physics-based differentiable rendering, aiming to improve 3D reconstruction quality.
Had the opportunity to showcase our reconstruction results at
(briefly featured at 1:13:20).
Our team’s work was accepted for publication at ICCV 2023
.
Education
Ph.D. Computer Science
University of California, Irvine
Supervised by
. Investigated the problem of physics-based inverse rendering, which aims to reconstruct the 3D shape and reflectance of an object from multiple images.