Vaibhav's Pages

Vaibhav Vaish

Ph.D Student
Computer Graphics Laboratory
Computer Science Department
Stanford University

Contact Info:

Computer Science Department
#360, Gates Building,
353, Serra Street,
Stanford, CA 94305.

Phone: 1-650-723-3154

Research

I'm working in computer vision and graphics at the Computer Graphics Laboratory. My advisor is Marc Levoy. My research interests span computational photography, 3D reconstruction from images, multi-view geometry, digital photo management. I defended my dissertation in November 2006 and officially graduated on 5th April 2007.

List of publications
Selected presentations

I am working with the Light Field Camera Array group on synthetic aperture imaging and reconstructing 3D shape using arrays of cameras and projectors. Here is an example illustrating synthetic aperture imaging:

Cameras aimed at bushes View from one camera of students behind bushes Synthetic aperture image

Synthetic Aperture Imaging. In this experiment, an array of 45 cameras was aimed at students hidden behind bushes (left image). It is almost impossible to make out the students from a single camera image (center). We can combine images from all 45 cameras to automatically construct a synthetic aperture image (right) in which the students behind the bushes are clearly visible. For more examples and details, please see [VV04], [VV05].

The ability to "see through" dense occlusions makes synthetic aperture imaging a potentially powerful tool for surveillance. Using simple projective geometry analysis, we have developed a simple and robust calibration procedure for synthetic aperture imaging, and a real-time system for tracking a person as he moves through a crowd. Synthetic aperture imaging allows us to simulaneously exploit both shape-from-stereo and shape-from-focus methods for 3D reconstruction. We have explored the space of 3D reconstruction methods enabled by a synthetic aperture in collaboration with Rick Szeliski's group at Microsoft Research [VV06].

We have created an archive of light fields acquired using our camera array and light field gantry for researchers to download and use in their work.

Older Projects

Nonlinear Minimization for Multi-Camera Calibration
Light Field Camera Calibration

Software

findgrid:
an automatic calibration grid detector.

Download linux binaries

Pygnome-games:
A traditional Indian game (untee, also called ashta chemma) and a modern brain teaser (brainvita). Implemented in python using the GNOME UI package and GTK2.

Download RPMs

Biographical Sketch

The highlight of my school days was participating in the Indian National Mathematics Olympiad, that left me with a lasting love of geometry. I graduated from Indian Institute of Technology, New Delhi, majoring in computer science and engineering. Those were four fantastic years, where I learnt cool stuff, made wonderful friends and had tremendous fun helping to maintain and administer the Intel Technology Laboratory.

I joined the PhD programme in computer science at Cornell in fall 2000, where I picked up ice hockey. After a year, I transferred to Stanford and graduated in April 2007.

Photos and Links

Stanford Photos

The Graphics Lab Ski Trip, 2002
Holi 2002
Diwali 2002
Siggraph 2003 2004: Summer in Seattle
2005: My parents' trip to USA

Memories from IIT Delhi

The Class of 2000
Our YearBook

All IITD Photos

If the images look dark, read our note about gamma correction.


Cameras aimed at bushes	View from one camera of students behind bushes	Synthetic aperture image
Synthetic Aperture Imaging. In this experiment, an array of 45 cameras was aimed at students hidden behind bushes (left image). It is almost impossible to make out the students from a single camera image (center). We can combine images from all 45 cameras to automatically construct a synthetic aperture image (right) in which the students behind the bushes are clearly visible. For more examples and details, please see [VV04], [VV05].