Top–down learning of low-level vision tasks
نویسندگان
چکیده
منابع مشابه
Top–down learning of low-level vision tasks
Perceptual tasks such as edge detection, image segmentation, lightness computation and estimation of three-dimensional structure are considered to be low-level or mid-level vision problems and are traditionally approached in a bottom-up, generic and hard-wired way. An alternative to this would be to take a top-down, object-class-specific and example-based approach. In this paper, we present a s...
متن کاملLearning Low-level Vision Learning Low-level Vision
We show a learning-based method for low-level vision problems{estimating scenes from images. We generate a synthetic world of scenes and their corresponding rendered images. We model that world with a Markov network, learning the network parameters from the examples. Bayesian belief propagation allows us to e ciently nd a local maximum of the posterior probability for the scene, given the image...
متن کاملUnifying Low-Level Vision
This white paper supports the goal of establishing Computer Vision as a coherent intellectual discipline by suggesting a specific agenda for the unification of many low-level vision principles, algorithms, and data structures. Our goal is to identify a set of highly related low-level vision problems, define their common structure, and establish a coherent intellectual discipline around the shar...
متن کاملLow Level Operations and Learning in Computer Vision
This thesis presents some concepts and methods for low level computer vision and learning, with object recognition as the primary application. An efficient method for detection of local rotational symmetries in images is presented. Rotational symmetries include circle patterns, star patterns, and certain high curvature patterns. The method for detection of these patterns is based on local momen...
متن کاملSynergism in Low Level Vision
Abstract Guiding image segmentation with edge information is an often employed strategy in low level computer vision. To improve the trade-off between the sensitivity of homogeneous region delineation and the oversegmentation of the image, we have incorporated a recently proposed edge magnitude/confidence map into a color image segmenter based on the mean shift procedure. The new method can rec...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Current Biology
سال: 1997
ISSN: 0960-9822
DOI: 10.1016/s0960-9822(06)00419-2