Matches in SemOpenAlex for { <https://semopenalex.org/work/W2097713019> ?p ?o ?g. }
Showing items 1 to 92 of
92
with 100 items per page.
- W2097713019 abstract "Many low-level vision algorithms assume a prior probability over images, and there has been great interest in trying to learn this prior from examples. Since images are very non Gaussian, high dimensional, continuous signals, learning their distribution presents a tremendous computational challenge. Perhaps the most successful recent algorithm is the Fields of Experts (FOE) [20] model which has shown impressive performance by modeling image statistics with a product of potentials defined on filter outputs. However, as in previous models of images based on filter outputs [30], calculating the probability of an image given the model requires evaluating an intractable partition function. This makes learning very slow (requires Monte-Carlo sampling at every step) and makes it virtually impossible to compare the likelihood of two different models. Given this computational difficulty, it is hard to say whether nonintu-itive features learned by such models represent a true property of natural images or an artifact of the approximations used during learning. In this paper we present (1) tractable lower and upper bounds on the partition function of models based on filter outputs and (2) efficient learning algorithms that do not require any sampling. Our results are based on recent results in machine learning that deal with Gaussian potentials. We extend these results to non-Gaussian potentials and derive a novel, basis rotation algorithm for approximating the maximum likelihood filters. Our results allow us to (1) rigorously compare the likelihood of different models and (2) calculate high likelihood models of natural image statistics in a matter of minutes. Applying our results to previous models shows that the nonintuitive features are not an artifact of the learning process but rather are capturing robust properties of natural images." @default.
- W2097713019 created "2016-06-24" @default.
- W2097713019 creator A5074429265 @default.
- W2097713019 creator A5079364492 @default.
- W2097713019 date "2007-06-01" @default.
- W2097713019 modified "2023-10-17" @default.
- W2097713019 title "What makes a good model of natural images?" @default.
- W2097713019 cites W2099068394 @default.
- W2097713019 cites W2101349782 @default.
- W2097713019 cites W2107790757 @default.
- W2097713019 cites W2123649031 @default.
- W2097713019 cites W2126159571 @default.
- W2097713019 cites W2130159466 @default.
- W2097713019 cites W2131686571 @default.
- W2097713019 cites W2136748901 @default.
- W2097713019 cites W2137234026 @default.
- W2097713019 cites W2140200866 @default.
- W2097713019 cites W2142615865 @default.
- W2097713019 cites W2145889472 @default.
- W2097713019 cites W2148852318 @default.
- W2097713019 cites W2153173219 @default.
- W2097713019 cites W4237434670 @default.
- W2097713019 cites W4237840503 @default.
- W2097713019 doi "https://doi.org/10.1109/cvpr.2007.383092" @default.
- W2097713019 hasPublicationYear "2007" @default.
- W2097713019 type Work @default.
- W2097713019 sameAs 2097713019 @default.
- W2097713019 citedByCount "263" @default.
- W2097713019 countsByYear W20977130192012 @default.
- W2097713019 countsByYear W20977130192013 @default.
- W2097713019 countsByYear W20977130192014 @default.
- W2097713019 countsByYear W20977130192015 @default.
- W2097713019 countsByYear W20977130192016 @default.
- W2097713019 countsByYear W20977130192017 @default.
- W2097713019 countsByYear W20977130192018 @default.
- W2097713019 countsByYear W20977130192019 @default.
- W2097713019 countsByYear W20977130192020 @default.
- W2097713019 countsByYear W20977130192021 @default.
- W2097713019 countsByYear W20977130192022 @default.
- W2097713019 countsByYear W20977130192023 @default.
- W2097713019 crossrefType "proceedings-article" @default.
- W2097713019 hasAuthorship W2097713019A5074429265 @default.
- W2097713019 hasAuthorship W2097713019A5079364492 @default.
- W2097713019 hasBestOaLocation W20977130192 @default.
- W2097713019 hasConcept C105795698 @default.
- W2097713019 hasConcept C106131492 @default.
- W2097713019 hasConcept C11413529 @default.
- W2097713019 hasConcept C119857082 @default.
- W2097713019 hasConcept C121332964 @default.
- W2097713019 hasConcept C154945302 @default.
- W2097713019 hasConcept C163716315 @default.
- W2097713019 hasConcept C167928553 @default.
- W2097713019 hasConcept C19499675 @default.
- W2097713019 hasConcept C31972630 @default.
- W2097713019 hasConcept C33923547 @default.
- W2097713019 hasConcept C41008148 @default.
- W2097713019 hasConcept C52740198 @default.
- W2097713019 hasConcept C62520636 @default.
- W2097713019 hasConcept C89106044 @default.
- W2097713019 hasConceptScore W2097713019C105795698 @default.
- W2097713019 hasConceptScore W2097713019C106131492 @default.
- W2097713019 hasConceptScore W2097713019C11413529 @default.
- W2097713019 hasConceptScore W2097713019C119857082 @default.
- W2097713019 hasConceptScore W2097713019C121332964 @default.
- W2097713019 hasConceptScore W2097713019C154945302 @default.
- W2097713019 hasConceptScore W2097713019C163716315 @default.
- W2097713019 hasConceptScore W2097713019C167928553 @default.
- W2097713019 hasConceptScore W2097713019C19499675 @default.
- W2097713019 hasConceptScore W2097713019C31972630 @default.
- W2097713019 hasConceptScore W2097713019C33923547 @default.
- W2097713019 hasConceptScore W2097713019C41008148 @default.
- W2097713019 hasConceptScore W2097713019C52740198 @default.
- W2097713019 hasConceptScore W2097713019C62520636 @default.
- W2097713019 hasConceptScore W2097713019C89106044 @default.
- W2097713019 hasLocation W20977130191 @default.
- W2097713019 hasLocation W20977130192 @default.
- W2097713019 hasOpenAccess W2097713019 @default.
- W2097713019 hasPrimaryLocation W20977130191 @default.
- W2097713019 hasRelatedWork W2134539662 @default.
- W2097713019 hasRelatedWork W2161803855 @default.
- W2097713019 hasRelatedWork W2254578859 @default.
- W2097713019 hasRelatedWork W2342588476 @default.
- W2097713019 hasRelatedWork W2350937085 @default.
- W2097713019 hasRelatedWork W2393870460 @default.
- W2097713019 hasRelatedWork W2899084033 @default.
- W2097713019 hasRelatedWork W3122849605 @default.
- W2097713019 hasRelatedWork W4284894156 @default.
- W2097713019 hasRelatedWork W3110774753 @default.
- W2097713019 isParatext "false" @default.
- W2097713019 isRetracted "false" @default.
- W2097713019 magId "2097713019" @default.
- W2097713019 workType "article" @default.