Raúl Díaz García

e-mail: rdiazgar (at) uci.edu
office: 4219 Donald Bren Hall

I was a PhD student at UC Irvine, where I was advised by professor Charless Fowlkes at the Computer Vision group.

My research focused on improving object detection by finding geometric context cues. In particular, I investigated how structure from motion and multi-view stereo can help in the world of scene understanding.

Since October 2016, I work at HP Inc., working on the Immersive Technology research group.

[resume] [Google Scholar] [LinkedIn]

PhD Thesis

R. Díaz, "Strong Geometric Context for Scene Understanding", UC Irvine, October 2016. [pdf]

Publications

R. Díaz, C. Fowlkes "Cluster-wise Ratio Tests for Fast Camera Localization", Technical Report , December 2016.
arXiv:1612.01689 [paper]

R. Díaz, M. Lee, J. Schubert, C. Fowlkes "Lifting GIS Maps into Strong Geometric Context for Scene Understanding", WACV , Lake Placid, NY, March 2016.
[paper]

R. Díaz*, S. Hallman*, C. Fowlkes, "Detecting Dynamic Objects with Multi-View Background Subtraction", ICCV , Sydney, Australia, December 2013.
[paper] [supplementary material]

R. Díaz*, S. Hallman*, C. Fowlkes, "Multi-View Background Subtraction for Object Detection", Scene Understanding Workshop , Portland, OR, June 2013.
[abstract] [poster]

* shared first author

Presentations

R. Díaz, "Strong Geometric Context for Image Understanding", Invited seminar talk, Computer Vision Center, UAB, July 2014. [video]