The availability of commercial light field cameras has spurred significant research into the use of light fields and multi-view imagery in computer vision. In this talk, we discuss our results over the past few years, focusing on a few themes. First, we describe our work on a unified formulation of shape from light field cameras, combining cues such as defocus, correspondence, and shading. Then, we go beyond photo-consistency, addressing non-Lambertian objects, occlusions, and material recognition. We also discuss applications for light field cameras such as motion deblurring and descattering. Finally, we show that advances in machine learning can be used to interpolate light fields from very sparse angular samples, in the limit a single 2D image, and create light field videos from sparse temporal samples.
|