Current
- Wenzel Jakob, Ph.D. CS
- Jonathan Kaldor, Ph.D. CS
- Kan Li
- Jon Moon, Ph.D. CS
Former
- Piti Irawan, Ph.D. CS
- Andrew Scukanec (now at Microsoft)
- Masato Ikura, BS CS (now at CMU ETC)
- Laura Beck, BS ECE
- Andrew Butts, BA CS (now at Pixar)
My research tends to focus on creating models for high-quality rendering. I'm particularly interested in accurate models for the appearance of everyday materials, including bidirectional reflectance and other models, and in acquiring and representing high-resolution geometry.
We are working toward realistic, physically based rendering of whole heads of hair by understanding and efficiently simulating the reflection of light, first from individual fibers and then as it reflects from fiber to fiber through large masses of hair.
This project with Henrik Wann Jensen and others went beyond the BRDF to study translucent materials, which allow light to scatter below the surface of an object. We developed an efficient approximation that works both for rendering and for measurement.
This major 3D scanning project digitized many of Michelangelo's large statues. I managed the continuing work, which involves digesting the vast collection of 3D and 2D data that was measured in Italy.
This project created a realistic model of a human face based on measured geometry, reflectance, texture, and motion.
We developed a new technique for measuring bidirectional reflectance using photographs of a curved sample. It is easy, inexpensive, quick, and accurate. We used it to make BRDF measurements of human skin.
Using a collection of calibrated photographs, this technique computed texture maps of intrinsic reflectance for a scanned 3D object. The results were realistic enough to match a photograph.
This work demonstrated inverse lighting and used the results for an experiment in altering the lighting in photographs of faces.
We expored the issues of aliasing in volume sampling and reconstruction, and we developed a metric for evaluating the quality of filters.