site best viewed in firefox Santosh Divvala

/home

/presentations

/code







Santosh Divvala

I am a Research Scientist at the Allen Institute for Artificial Intelligence (AI2) led by Oren Etzioni. I also collaborate with Ali Farhadi and Carlos Guestrin at the University of Washington. I received a PhD in Robotics from Carnegie Mellon University, where I worked under the supervision of Martial Hebert and Alyosha Efros.

My interests are in Computer Vision, specifically the Image Understanding problem. My approach to this problem takes a Machine Learning perspective, leveraging today's Internet-scale image and video databases. The results of my research apply to Robotics (building systems that can aid us in better interpreting and reasoning about the world) & to Computational Photography (enhancing the way we relish our photo collections).



For full-list of publications, please see my Google Scholar Page







   

Selected Projects

Learning Everything about Anything

with Carlos Guestrin, Ali Farhadi

How important are 'Deformable Parts' in the Deformable Parts Model?

with Alyosha Efros, Martial Hebert


Object Instance Sharing by Enhanced Bounding Box Correspondence

with Alyosha Efros, Martial Hebert


Empirical Study of Context

with Derek Hoiem, James Hays, Alyosha Efros, Martial Hebert


Annotating Google Streetview

with Mei Han, Sergey Ioffe

Web-scale Scene Parsing

with Svetlana Lazebnik, Alyosha Efros, Martial Hebert


Unsupervised Poselets

with Larry Zitnick, Ashish Kapoor, Simon Baker


Space-Carving Approach to Surface Estimation

with Drew Bagnell, Martial Hebert


Autonomous Image-based Exploration for Mobile Robot Navigation

with C V Jawahar


Visual Servoing in Unconventional Environments

with C V Jawahar