News Highlight

Title: Benchmarking Stereo Vision and Optical Flow Algorithms

Speaker: Daniel Scharstein (http://www.cs.middlebury.edu/~schar/), Middlebury College

Time: Friday, January 25th, 4:00pm-5:00pm

Location: Babbio Center 220

Host: Philippos Mordohai


Abstract:

Stereo vision and optical flow methods attempt to measure scene depth and motion by matching and tracking pixels across images.  To evaluate the performance of such methods, we need "ground truth" -- the true depth or true object motion.  In this talk I will describe different techniques for creating image datasets with ground truth, including structured lighting, laser and CT scanners, and hidden fluorescent texture.  The Middlebury datasets, created in collaboration with undergraduates, are now well-established benchmarks in computer vision, and I will discuss both benefits and potential pitfalls of such benchmarks.  I will also briefly touch on how data with ground truth can aid in developing new algorithms.

Bio:
Daniel Scharstein, Professor of Computer Science at Middlebury College in Vermont, studied Computer Science at the Universitaet Karlsruhe, Germany, and received his PhD from Cornell University in 1997.  His research interests include computer vision, image-based rendering, and robotics.  He maintains several online computer vision benchmarks at http://vision.middlebury.edu.