Yuriy Vasilyev

Yuriy Vasilyev

PhD. Candidate

Harvard University
School of Engineering and Applied Sciences

Brief Resume

Contact Information:

Maxwell Dworkin 215
33 Oxford Street
Cambridge, MA 02138

email: lastname@fas.harvard.edu

phone: (617)495-9760

Research and Publications

Working with my advisor Dr. Todd Zickler, my main research interest is shape inference of mirror-like objects.

Toward a Theory of Shape from Specular Flow
Yair Adato, Yuriy Vasilyev, Ohad Ben-Shahar, and Todd Zickler.
ICCV 2007 Oral. Paper [pdf] Oral Presentation Slides and Notes [66MB ppt]
Relative motion between a curved, specular (mirror-like) surface and its environment induces a motion field on the image plane--termed specular flow--that provides direct access to surface shape information. Here we derive a shape-from-specular-flow equation and examine a special class of motions for which shape can be recovered in closed form.

Dense Specular Shape from Multiple Specular Flows
Yuriy Vasilyev, Yair Adato, Todd Zickler, and Ohad Ben-Shahar.
CVPR 2008. Paper [pdf] Poster [3MB pdf]
Specular flows due to multiple unknown motions can be combined linearly to yield the special case described above. Moreover, the coefficients of this linear combination can be determined directly from the observed flows. The result is an "auto-calibrating" system for the recovery of shape from arbitrary specular flows.

Dense Specular Shape from Multiple Specular Flows
Fabiano Romeiro, Yuriy Vasilyev, Todd Zickler.
ECCV 2008 Oral. Paper [pdf]
Existing systems for measuring reflectance are cumbersome, and although the process can be streamlined using cameras, projectors and clever catadioptrics, it generally requires complex infrastructure. We propose a simpler method for inferring reflectance from images, one that eliminates the need for active lighting and exploits natural illumination instead.

Teaching

In the Spring 2007 and Spring 2008 semesters I was a Teaching Fellow for QR 20 - Computers and Computing, an introductory programming course taught in python to students concentrating in humanities and liberal arts.

	Toward a Theory of Shape from Specular Flow Yair Adato, Yuriy Vasilyev, Ohad Ben-Shahar, and Todd Zickler. ICCV 2007 Oral. Paper [pdf] Oral Presentation Slides and Notes [66MB ppt] Relative motion between a curved, specular (mirror-like) surface and its environment induces a motion field on the image plane--termed specular flow--that provides direct access to surface shape information. Here we derive a shape-from-specular-flow equation and examine a special class of motions for which shape can be recovered in closed form.
	Dense Specular Shape from Multiple Specular Flows Yuriy Vasilyev, Yair Adato, Todd Zickler, and Ohad Ben-Shahar. CVPR 2008. Paper [pdf] Poster [3MB pdf] Specular flows due to multiple unknown motions can be combined linearly to yield the special case described above. Moreover, the coefficients of this linear combination can be determined directly from the observed flows. The result is an "auto-calibrating" system for the recovery of shape from arbitrary specular flows.
	Dense Specular Shape from Multiple Specular Flows Fabiano Romeiro, Yuriy Vasilyev, Todd Zickler. ECCV 2008 Oral. Paper [pdf] Existing systems for measuring reflectance are cumbersome, and although the process can be streamlined using cameras, projectors and clever catadioptrics, it generally requires complex infrastructure. We propose a simpler method for inferring reflectance from images, one that eliminates the need for active lighting and exploits natural illumination instead.