Vasiliy Karasev

I am a senior software engineer at Waymo where I work on computer vision for autonomous driving.

Previously, I spent several years doing computer vision at Zoox. I did my PhD at UCLA, and was advised by Prof. Stefano Soatto. My work revolved around ''value of information'' / decision making problems, optimization, and their applications in computer vision.

During my studies I had the opportunity to visit Honda Research Institute and Samsung Display. I received my bachelors from UC Berkeley.

     

profile photo

Publications

project image

Intent-Aware Long-Term Prediction of Pedestrian Motion


V. Karasev, A. Ayvaci, B. Heisele, and S. Soatto
ICRA, 2016
video / details / pdf / bibtex

Forecasting what pedestrians intend to do is easier if they behave rationally. We show how this assumption simplifies motion prediction in the assisted/autonomous driving setting.

project image

Causal video object segmentation from persistence of occlusions


B. Taylor, V. Karasev, and S. Soatto
CVPR, 2015 (oral)
video / details / pdf / bibtex

We show how to exploit occlusions to discover salient objects in video.

project image

Active frame, location, and detector selection for automated and manual video annotation


V. Karasev, A. Ravichandran, and S. Soatto
CVPR, 2014
video / details / pdf / bibtex

How to choose where and which detectors to run (or m-turks to query), if we can run only a few of them? We answer this question using the ‘‘information gathering’’ framework, and show results on semantic video segmentation.

project image

Controlled recognition bounds for visual learning and exploration


V. Karasev, A. Chiuso, and S. Soatto
NIPS, 2012
pdf / bibtex

We show how to (greedily) search for an unknown object, under occlusions, quantization-scale, and uncertain measurements.

project image

Temporal presentation protocols in stereoscopic displays: flicker visibility, perceived motion, and perceived depth


D. Hoffman, V. Karasev, and M. Banks
Journal of the Society for Information Display, 2011
pdf / bibtex

We studied how different 3D display presentation methods affect flicker, motion artifacts, and errors in perceived depth.

project image

High resolution fMRI using compressed sensing


2011
details / pdf

This was my MS project.
Compressed sensing in dynamic MRI is normally used to improve temporal resolution. Here we used it to improve spatial resolution. Reconstruction used the simplest possible TV regularization.




Design and source code from Leonid Keselman's adaptation of Jon Barron's website